Documente Academic
Documente Profesional
Documente Cultură
TRANSMISSION AND
SUBSTATION DESIGN
AND OPERATION
Since 1968 SYMPOSIUM
2017 TECHNICAL PAPERS
TABLE OF CONTENTS
SESSION TITLE AUTHOR(S) PAGE
Utilizing the Design of Hybrid Transmission Jessica McLarty, P.E., James R. Childs, P.E.,
165 - 177
Line Structures in Flood Plain Regions John L. Webb, P.E.
Visit www.tsdos.org for additional conference materials, including the speaker slides, presented at the 2017 TSDOS.
Additional 50th Annual TSDOS technical papers and the conference presentation slides are available at www.tsdos.org.
Technical papers available on flash drive sponsored by:
Entergy Services, Inc. Hartburg Substation Flooding and Recovery
July 2017
Prepared for:
The 50th Annual
Transmission and Substation Design and Operation Symposium
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
Background
In 2015, Black & Veatch was retained by Entergy to provide full turn-key Engineering,
Procurement, and Construction (EPC) services for the expansion of their Hartburg substation
located in southeast Texas, near the city of Deweyville, along the Sabine River. The project
involved the addition of a new 500/230 kV auto-transformer and the build out of the 230 kV yard
into a four- position ring. The construction of the project was nearing completion by early 2016,
with a target in-service of July. As the construction crews began the final push in early March,
south Texas came under one of the heaviest rainfalls in recent times.
Some 60 miles north of the Hartburg substation, along the Sabine River is the Toledo Bend
Reservoir. Managed by the Sabine River Authority, the Reservoir was completed in 1969 to
provide a number of public benefits, including water supply, recreation and hydropower
production. It was never intended to be a flood control facility for the Sabine River. It covers over
180,000 acres, making it the fifth-largest man-made lake in the United States.
Hartburg substation is situated near the Sabine River, downstream of the Toledo Bend Reservoir..
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
The Flood
By spring 2016, the area had seen a steady return of rainfall after several years of severe drought,
which had brought the reservoir to near capacity. The heavy rains started on March 9 and dumped
18 to 24 inches of rain on the drainage basin for the upper Sabine River. With the water level
rising above record levels and threatening the integrity of the dam and spillways, the Sabine River
Authority opened the flood gates to substantially reduce the reservoir water level. During the
peak flow period on and around March 10, the flood gates released 1.5 million gallons of water
per second to flow down river- a flow rate that is more than double that of Niagara Falls. The
resulting deluge of water was expected to cause flooding in the lower Sabine River of two to six
feet.
Between March 11 and March 13, noticeable floodwaters started to affect the Hartburg
substation, which sits just two miles west of the main channel of the Sabine River and is at the
edge of inhabitable land. During this time Entergy’s crew and the Black & Veatch construction
crew (Overland Contracting, Inc.) did as much pre-flood preparation as possible to mitigate flood
damage. For example, Entergy’s crew installed temporary flood walls around each of the three
control enclosures to prevent floodwaters from seeping inside. Two of these control enclosures
existed prior to commencement of the project. The third control enclosure was installed as part of
this project and will be discussed in further detail later in this paper. Overland Contracting, Inc.
(OCI) relocated uninstalled equipment out of the substation and secured much of the loose
material that might float and cause damage during the flood. Coupling capacitor voltage
transformers (CCVT) that were still crated and yet to be installed were removed. The remainder
of the major equipment was already installed prior to the flood event.
On March 14, the floodwaters rolled in and submerged the substation under five to seven feet of
water. By the time the local residents got to safety, the waters were over nine feet above the 24-
foot flood stage. The Sabine River at Deweyville crested to an historical level of 33.24 feet on
March 15.
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
Damage Assessment
For a week, the wide spread of the Sabine River covered the station. On March 21, the
floodwaters had subsided enough to safely gain access to the Hartburg substation and begin the
damage assessment. Entergy, Black & Veatch, and OCI were on site to perform a full walk down
to begin the process of determining the extent of the damage caused by the flood. Both the 500kV
yard and the 230kV yard had sustained flooding and subsequent damage. Since the original scope
of work of the project was limited to the 230kV yard, new autotransformer, and one 500kV
breaker and associated switches, Black & Veatch performed the restoration work in the 230kV
yard while coordinating closely with Entergy, who performed restoration for the remainder of the
existing 500kV yard. The following major equipment was damaged during the flood:
The flood also damaged or swept away miscellaneous bus hardware, circuit breaker installation
material, cable trench lids, and transmission line hardware that crews did not have time to move
to safety.
Recovery Efforts
Shortly after the damage was assessed, Black & Veatch and Entergy worked together to come up
with a recovery plan, which would include the repair or replacement of equipment and the raising
of the major 230kV equipment. Some of the decisions to re-use or replace flood-affected material
were based on the recommendation of the manufacturers. However, the decision on how high to
raise the equipment required more review. The scope of the event was considered a 500-year
flood event, which set a clear flood stage level for the area. Given the known water line in most
of the control cabinets, Entergy decided on a net increase of six feet; from a base elevation of 21
feet to a new top-of-concrete (TOC) of 27 feet. The net increase of six feet included the known
high water mark of 24’, which represented three feet of the six feet total, and an additional three
feet to account for the projected peak of the flood event. The projected peak was provided by the
Sabine River Authority during their decision-making process to release water from Toledo Bend.
The following is a detailed description of the major equipment recovery and elevation changes:
The 230kV control enclosure was installed as part of this project’s scope of work. The control
enclosure had already arrived on site, been set on piers, and installed per manufacturer’s
recommendations. Most of the control cable had been pulled between the control enclosure and
the 230kV equipment. Crews were working to complete wire terminations as news of the
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
projected flood event was relayed to the team. To mitigate damage to the installed equipment,
Entergy installed a temporary flood wall around the control enclosure.
Although there was a temporary flood wall around the 230kV control enclosure, two feet of water
penetrated the control enclosure and caused extensive damage to the flooring and the walls. The
control enclosure manufacturer and relay panel manufacturer sent an engineer to the site to assess
the damage from the flood and develop a plan for restoring the enclosure and equipment to pre-
flood conditions. Over five feet of the interior walls had to be gutted because of water damage to
the gypsum board. The floors had to be removed to enable replacement of the water-damaged
gypsum board and spray polyurethane foam. Any equipment inside of the control enclosure that
was submerged in floodwater was replaced. The manufacturer of the relay panels recommended
all the control/relay panels be removed and returned to their facility for testing, cleanup, and
repair or replacement because of their exposure to the humidity, floodwater contaminates, and
high temperatures for over five days. OCI removed the panels and equipment, sent them to the
vendor for refurbishment, and re-installed and re-wired all the equipment upon its return. In
addition, OCI refurbished the building including the gypsum board and the floor, following
recommendations of the control enclosure manufacturer.
During the refurbishment, Black & Veatch & Entergy worked on a plan to raise the control
enclosure in a timely manner. Entergy requested that the top of concrete elevation for the control
enclosure be increased to 28 feet from the original design of 21 feet. The decision was made to
pour new drilled piers next to the existing location. This would limit the control enclosure to
being moved a single time and allow the refurbishment process of the control enclosure to
coincide with the installation of the new piers. The control enclosure manufacturer then
confirmed that the enclosure would be within its design parameters, even though it was going to
be installed seven feet above grade.
Once the control enclosure was completely refurbished and the relay/control panels re-installed,
Entergy and Black & Veatch worked with the building manufacturer on a lifting plan. The control
enclosure was originally shipped to the site in two halves and bolted together. Each half was 42
feet by 10 feet by 12 feet 2 inches tall, and weighed 45,000 pounds. This resulted in an installed
control enclosure that was 42 feet by 20 feet by 12 feet 2 inches tall, weighing 90,000 pounds.
The overall size of the control enclosure and the split design offered a unique set of challenges for
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
the lift. The lifting company was not able to use typical lifting beams, so they designed and
fabricated beams that could provide the necessary support to complete the lift. Multiple meetings
were held to discuss the lifting plan and to grant all stakeholders an opportunity to discuss
potential issues. A third party specializing in crane rigging and lift plans was brought in by
Entergy to review the lift plan to be certain that every potential issue was accounted for in the
final plan. Once Entergy, Black & Veatch, OCI, and the control enclosure manufacturer agreed
that the lifting plan mitigated all concerns with moving the control enclosure, the lift was
scheduled.
The photos below illustrate the lift. As can be seen, the building did not need to be moved far.
However, the split design of the building and the need for a tandem crane lift complicated the
move. The lift and relocation of the control enclosure went smoothly. The importance of having a
detailed lift plan as described above was crucial to the success of the lift.
230kV Breakers
Even though the floodwaters reached the bottom of the 230kV breaker mechanism cabinets, it did
not reach the equipment contained within. However, Entergy decided to raise the breakers to
ensure any future flood event would not damage the breaker mechanism cabinet. The minimum
elevation of the bottom of the breaker cabinet was set at 27 feet as discussed in the recovery
efforts section. Black & Veatch worked with the breaker vendor to determine the best option to
raise the breakers. The breaker vendor had a design for a breaker with taller legs. Fortunately,
these extended height legs would raise the bottom elevation of the breaker mechanism cabinet to
exceed the minimum required elevation of 27 feet. And perhaps as important, to minimize further
delays in the scheduled in-service, the vendor had leg extensions readily available on short notice.
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
The autotransformer current transformer (CT) junction box and 500kV and 230kV CCVT
junction boxes were damaged during the flood event. There was evidence that the floodwaters
had entered the junction boxes. The decision was made to raise the bottom of the junction box
elevation to 27 feet. This height was established based on the same criteria stated in the recovery
efforts section. For the autotransformer CT junction box, this was accomplished by installing a
new stand for the junction box. The 500kV CCVT junction boxes were remounted at a higher
elevation on the CCVT stands. The 230kV CCVT stands were replaced with taller stands to allow
for mounting the junction boxes at the proper heights.
The 500kV and 230kV switch motor operators were submerged during the flood and damaged.
All operators needed to be replaced. The decision was made to raise the motor operators to meet
the new 27 foot minimum elevation to bottom of equipment. It will be discussed later that the
500kV motor operators were moved to an elevation of 26 feet.
Raising the 230kV motor operators posed some unique challenges. The operators were mounted
on the leg of H-frame dead-end structures. The first thing that needed to be considered was
whether there would be clearance issues for the motor operator when mounted at a higher
elevation. There was much coordination between the substation packager and the switch vendor
to confirm that once the motor operators were raised they would be able to operate as intended.
To complete the installation, new mounting brackets were fabricated for the motor operators. The
switch vendor provided a field service representative on-site to mount new brackets and make
field adjustments as necessary. Once this was complete, OCI was able to install the new motor
operators. Installers discovered that the decreased length of pipe on the switch operator meant
that there was less room to mount some of the switch accessories. The switch status indicator
needed to be relocated to a different part of the operating mechanism as there was no longer room
for it to be installed as originally designed.
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
The station service voltage transformer (SSVT) disconnect switch was damaged during the flood
event. The decision was made to raise it to the 27 foot minimum elevation using the criteria set
forth previously. There was discussion on whether or not to raise the SSVT. The vendor did not
have a design available for a taller stand and would not honor their seismic guarantees if the
existing stand was modified. Through a closer look at the SSVT it was determined that the
secondary cabinet on the SSVT was currently above the 27 foot minimum elevation. The SSVT
could remain at its present elevation. The disconnect switch was raised by replacing the existing
stand with a taller one.
The 500kV breaker is an independent pole operated/operators (IPO) design and the marshalling
cabinet had a few inches of water inside. The decision was made to raise the cabinet to the 27 foot
minimum elevation using the criteria set forth previously. This presented another set of
challenges. Raising the marshalling cabinet resulted in the need for new cables to be supplied by
the vendor. The original cabinet location was located near the middle phase breaker. Raising it in
this location would create a phase-to-ground clearance violation between the cabinet and either of
the two other phases. There was not a location within the breaker footprint that allowed for the
cabinet to be raised to the required elevation. Ultimately, the cabinet was moved outside the bus
and breaker footprint by 20 feet and raised to meet the required minimum elevation. A new stand
was designed and installed to achieve the desired elevation.
There were several inches of water in the autotransformer control cabinets. The transformer
vendor determined it would take a significant amount of work to raise the cabinets on the
transformer in the field. The required amount the bottom of the cabinets needed to be raised was 3
feet. The transformer vendor recommended that the entire transformer be raised 3 feet to mitigate
future flooding issues with the control cabinets. Entergy’s field personnel were consulted, and it
was decided to leave the transformer cabinets and the transformer at their current elevation. Part
of this decision was based on the effort involved with moving three single-phase 500/230kV
autotransformers weighing over 330,000 pounds each. Each unit would also need over 12,000
gallons of oil removed and refilled after completion of the raised installation.
Most of the control cables necessary for completion of the project had been cut and pulled at the
time of the flood. OCI was in the process of terminating conductors at the panels and control
cabinets when pre-flood preparation began. Cable that was not terminated was left in the cable
trench, and cable that remained on the reel was left during the flood. Black & Veatch and OCI
worked with the cable manufacturer to determine if the cables on site during the flood could be
used post flood. Due to the fact that the ends of the cables were not sealed, it was determined by
the manufacturer that the cables could not be reused since it was possible that water had entered
the cable jacket.
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
Once the decision was made to raise and move the control enclosure and raise the control cabinets
in the yard, much of the existing cable would not have been available for reuse as the cut lengths
would be too short to reach the new connection locations.
Maintenance Platforms
Entergy requested that permanent maintenance platforms be installed for all equipment and
cabinets that were raised. This was requested so that maintenance, testing, and future construction
personnel would not need to perform work at the equipment while standing on ladders. Black &
Veatch worked with Entergy to establish criteria for the sizing and height of the maintenance
platforms. The criteria were based on providing field crews ample space to perform job tasks
safely and to allow maintenance work to be accomplished at a suitable height.
The following sizes and heights were determined to be appropriate for this application:
500kV Motor operated switch (MOS) – 6-feet x 10-feet platform, 3 feet 3 inch height
500kV Breaker Cabinet – 5-feet x 6-feet platform, 4 feet 3 inch height
500kV CCVT – 5-feet x 5-feet platform, 3 feet 3 inch height
230kV Equipment – 5-feet x 5-feet platform, 4 feet 3 inch height
The final heights for platforms were selected to minimize the number of unique platform
installations. Black & Veatch started with the preferred heights supplied by Entergy and
performed a National Electrical Safety Code (NESC) 124 clearance analysis. Platform heights
were adjusted as needed so that clearances could be met without the need for installation of
equipment safety guard rails. In the 230kV yard, there were clearance issues between the MOS
platforms and the installed surge arresters. Spacers were added to the surge arrester stands to
allow for the preferred MOS platform height.
Ultimately, it was determined that one height could be used for all 230kV equipment and thus one
platform design was utilized. Each piece of 500kV equipment had different workspace area needs
and resulted in three different designs.
Some consideration was given to shipping the platforms to the site pre-assembled. This would
reduce the amount of time required for installation; however, it would increase shipping costs due
to a higher number of trucks being required to complete the delivery. Ultimately, it was decided
to ship the platforms broken down and assemble on site.
Black & Veatch also designed and installed new stair platforms for the main entrance and the two
battery room entrances for the control enclosure. Any instances that led to equipment needing to
be turned 90 degrees to enter a doorway from the top of the stairs required a minimum of an
eight-feet by eight-feet platform. Instances that allowed for equipment to enter a doorway straight
from the stairs required a minimum of six-feet by six-feet platform. These criteria were used to
size the final platforms for the main control enclosure doors and the two battery room doors.
Lessons Learned
The successful outcome of these efforts exemplifies the team’s commitment to communication.
Black & Veatch, OCI, and Entergy held twice-weekly planning meetings in the weeks leading up
to the flood event and in the months after. The meetings were used to discuss mitigation plans and
to prepare for restoration activities. As needed, vendors were brought into the meetings to provide
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
status updates on deliveries. These meetings were crucial to the success of the project. By
meeting twice a week, all stakeholders were kept up to date on the status of all deliverables. It
also helped to determine which tasks were the critical paths.
Another lesson is to always take the time to step back and see the big picture. During the flood
mitigation and restoration efforts, it was determined that the backup station service padmount
transformers needed to be raised to the 27 foot minimum elevation. After some discussion
between Black & Veatch and Entergy, it was decided to keep this requirement in the final design.
The padmount transformer had already been purchased but had not arrived on site prior to the
flood event. Since the padmount had been purchased already, the design team looked into options
of building a stand for the padmount or pouring a pier with seven feet exposed similar to the
control buildings. It was ultimately determined to pour a pier with the seven foot reveal.
The station service transformer was raised 7 feet above grade with a massive pier foundation.
This may be the first and only pier mounted padmount-design station service transformer. After
seeing photos of the completed installation, a better approach may have been to purchase a pole
top transformer, install it on a wood pole in the same location, and mount the safety disconnect
switch to the pole. This would have been a less substantial installation, eliminating a large pier
and stand from the installation requirements.
Conclusion
Throughout the process, Black & Veatch worked closely with Entergy, thoroughly documenting
and reporting project activities while maintaining close communication with Entergy
management. This enabled Entergy to provide information to their internal organization and to
other stakeholders in a timely manner.
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
The flood restoration and project adjustments took substantial collaboration between Black &
Veatch, Entergy, and multiple vendors. The unique design challenges required creative, efficient
solutions by the team. Despite the significant setbacks of the record-setting flood, the project was
only four weeks behind the original in-service schedule because of the responsive and proactive
mitigation efforts of the team.
© Black & Veatch Holding Company 2017. All Rights Reserved. The Black & Veatch name and logo are registered
trademarks of Black & Veatch Holding Company.
Submitted To
50th Annual Transmission and Substation
Design and Operation Symposium (TSDOS)
of
The University of Texas at Arlington
Page 15 of 284
The Catenary Exposed
Understanding Theoretical Conductor Behavior in Transmission Lines
ABSTRACT
The catenary describes the most unique and important shape in transmission engineering—the
hanging conductor. The loads imposed by conductors under various temperature, wind, and ice loading
conditions drive transmission line structural design. Fundamentals of catenary mechanics, including the
equations of position, tension, and elongation, are presented herein in a manner that is concise and easy to
understand. These catenary formulas can be used in simple spreadsheets to obtain results that agree with
sophisticated computer software packages for typical transmission line engineering problems. Recent
reliability and uprating work driven primarily by the North American Electric Reliability Corporation
(NERC), coupled with a retiring work force, have led to a significant influx of young engineers in the power
delivery industry over the last decade. Not surprisingly, many new engineers have quickly learned the trade
and become expert users of industry standard computer software. However, it is questionable how
effectively the knowledgebase of fundamental hand-calculation methods—used for over a century—are
being transferred to these young engineers. This practice, if not well-managed, may have unforeseen
consequences on various aspects of projects including budget, schedule, and safety of the design. This paper
aims to expose otherwise intimidating catenary formulas and to equip transmission engineers of all ages
with a deeper understanding of theoretical conductor behavior and useful hand-calculation methods.
Keywords: catenary, cable, transmission line
INTRODUCTION
The hanging cable is a fascinating structural member. While complex in regards to large
displacements, material nonlinearities, bimetallic or composite materials, and flexible boundary supports,
a small cable element is quite simple as loads are transferred via tension only and not bending. The hanging
cable deforms in the shape of a “catenary”—a word derived from Latin meaning “chain.” The catenary
describes a hanging wire or chain supported at ends under a uniform load. Mathematically, the catenary
shape is a function of the hyperbolic cosine. The catenary shape appears in many forms in nature including
a hanging chain, spider web, necklace, rope in moving water (supported at ends), etc. The catenary equation
was derived independently by mathematicians Leibniz, Huygens, and Johann Bernoulli in 1691 in response
to a challenge by Jakob Bernoulli (Weisstein 1999). Many relationships can be easily derived from the
basic catenary equation of position including slope, tension, length, sag, etc.
Page 16 of 284
While of paramount importance in transmission engineering, interestingly, few practicing
engineers seem to have taken time to become intimately acquainted with the catenary equation—probably
due, in part, to the perceived complexity. To illustrate, transmission engineers have long known that the
hanging conductor is a catenary, rather than a parabola, but have often used less computationally expensive
parabolic approximations since the two shapes are similar when curvature is small. In recent decades, with
the advent of the computer, the actual catenary equations (as opposed to parabolic approximations) are now
commonly utilized to compute tensions in software, but it appears that today a true working knowledge of
these equations is limited primarily to software engineers and academics. This should not be the case—
transmission engineering practitioners should also be well versed with the catenary equations. The shape
of the hanging conductor under uniform load is complex, but not complicated to understand.
Additionally, many fundamental catenary equations, though well published, are unfortunately not
readily accessible in compiled form, particularly for the beginning engineer. Today’s practicing
transmission engineer is often exposed to computer software so early that many fundamental theory and
hand-calculation methods are overlooked. The “sweet spot” of structural design and analysis is achieved
when using a combination of experimental and/or field data, mechanics analysis (i.e. hand-calculations),
and numerical analysis (e.g. finite element software) to determine results (Figure 1). Without all three of
these components, a structural design or analysis may appear sound, but could be off mark due to
complexities of actual structural behavior.
The purpose of this paper is to expose otherwise intimidating catenary formulas and to equip
transmission engineers with a deeper understanding of theoretical conductor behavior. Basic catenary
formulas are shared which can be easily programmed into simple spreadsheets to obtain results that agree
with sophisticated computer software packages for single-span problems. With a better grasp of
fundamental equations, engineers will be empowered to verify complex designs created using sophisticated
software tools, such as PLS-CADD® (1993), with simple hand-calculation methods.
Mechanics Numerical
Analysis Analysis
(i.e. Hand- (e.g. Finite
Calcs) Element)
Page 17 of 284
RIGID BODY BEHAVIOR (STATICS)
The hanging cable may be considered a rigid body once at rest (Figure 2). The cable deforms so
that internal forces result in no shear or bending moment, only tension. To be a catenary, a member must
assume the following characteristics based on geometry, material properties, and boundary conditions:
1. Simply supported at ends;
2. Uniform force per unit length; and,
3. Perfectly flexible.
Note that for the rigid body discussion the member is also assumed inextensible.
y(x) = a * cosh(x/a) 2
h 3
w s
1
L
λ1 λ2
λ
a
The catenary shown in Figure 2 is asymmetric. This is important as uneven end points are quite
typical in real-world applications. The symmetric catenary may be treated as a special case, but is very
useful for formulating approximate hand-calculation methods since symmetry simplifies the equations
greatly. As a side note, the well-known Ruling Span method (Thayer 1924) utilizes symmetry to determine
a stringing tension by idealizing a multi-span section as a single, weighted-average “ruling” span.
The discussion of the catenary begins with stress boundary conditions which include simple
supports at the ends and uniform force per unit length. The uniform force per unit length may be due to
either internal body forces (self-weight) or externally applied forces of surrounding fluid (wind, water, etc.).
The cable force per unit length, w, with cable material density, ρ, gravitational acceleration, g, and cable
cross-sectional area, A, is defined as:
𝑤 = 𝜌𝑔𝐴 (1)
Page 18 of 284
The member must also be perfectly flexible, which implies zero bending stiffness. To elaborate,
consider the difference between a simply-supported beam and a simply-supported cable, each supporting
self-weight only. The beam is highly rigid and supports self-weight by transferring shear via bending,
resulting in small deformations, forming the shape of a parabola. In contrast, the cable is highly flexible
and supports self-weight by transferring tension only, resulting in large deformations, forming the shape of
a catenary. Interestingly, a cable may be regarded as simply a very slender beam. Thus, a transition point
must exist in which an increasingly slender beam loses bending stiffness and begins to transfer tension
rather than shear, becoming a cable.
A derivation of the catenary equation is published by Hibbeler (2007). The equation of position in
2-dimensional Cartesian coordinate system is given in terms of the catenary constant, a:
𝑎
𝑦(𝑥) = 2 (𝑒 𝑥⁄𝑎 + 𝑒 −𝑥⁄𝑎 ) = 𝑎 ∗ 𝑐𝑜𝑠ℎ(𝑥/𝑎) (2)
It is important to note that most transmissions engineers are familiar with the basic catenary
equation of position. However, three main reasons why transmission engineers may likely be intimidated
by this equation are: 1) unfamiliarity with hyperbolic functions; 2) catenary constant must be solved
iteratively (not directly); and, 3) confusion regarding transformation from local to global coordinate systems
(e.g. transformation to station-elevation coordinates). This remainder of this section provides the engineer
instruction regarding these three concerns, indicates useful applications of the catenary used in computer
software, and introduces additional useful catenary equations—which may be derived from the basic
equation of position, including slope, arc length, sag, and tension.
1) Unfamiliarity with hyperbolic functions. Hyperbolic functions are simple and analogous to
trigonometric functions (very similar identities). However, in contrast to trigonometric functions which deal
with angles, hyperbolic functions deal in terms of Euler’s number, e = 2.718… The hyperbolic cosine
function is less intimidating when broken down in terms of this number. For example, Figure 3 shows that
the most basic catenary equation is simply a sum of ex and e-x, multiplied by ½. The figure depicts each of
these components, as well as the catenary equation with a constant equal to 1 (solid line).
For convenience, the definitions of the hyperbolic cosine and sine functions are provided. These
definitions will be useful for the remainder of the discussion:
1
𝑐𝑜𝑠ℎ(𝑢) = 2 (𝑒 𝑢 + 𝑒 −𝑢 ) (3a)
1
𝑠𝑖𝑛ℎ(𝑢) = 2 (𝑒 𝑢 − 𝑒 −𝑢 ) (3b)
2) Catenary constant must be solved iteratively. As far as solving for the catenary constant,
which cannot be solved directly, engineers designing a spreadsheet may take advantage of built-in
capabilities in common software, such as Excel’s Goal Seek functionality, to provide solutions. More
Page 19 of 284
ambitious engineers may wish to write custom root-finding algorithms using iterative techniques such as
Newton-Raphson.
Axis Title
4
y
y = ex + e-x
3
2
y = 1/2*(ex + e-x)
= cosh(x)
1
y = ex y = e-x
© 2017 D. Gelder
0
-4 -3 -2 -1 0 1 2 3 4
x
3) Confusion regarding transformation from local to global coordinate systems. Another major
source of confusion is in regards to coordinate system transformation. The catenary equation is formulated
in terms of a local coordinate system located a distance, a, below the low point of the curve. However,
engineers typically use a global coordinate system relative to the transmission line alignment (e.g. station-
elevation) for drawings, calculations, etc. Thus, the local coordinate origin is typically back-calculated
based on some combination of known global coordinates and possibly tension or wire length. Note that for
all asymmetric catenary curves (most cases) as tension changes the local coordinate system becomes a
moving target relative to the global system. In other words, the low point of the curve moves relative to
fixed end points defined in global coordinates.
Transformation of the catenary equation from local (x,y) to global coordinates (X,Y), located a
distance (Δx,Δy) from the local coordinate system is demonstrated:
𝑋−∆𝑥
𝑌 − ∆𝑦 = 𝑎 ∗ 𝑐𝑜𝑠ℎ ( ) (4)
𝑎
where,
𝑋 = 𝑥 + ∆𝑥
{ (5)
𝑌 = 𝑦 + ∆𝑦
Page 20 of 284
A unique catenary oriented relative to a plane can be defined by three discrete points (Figure 2).
Therefore, the catenary equation may be transformed from local to global coordinates using a system of
equations which can be used to solve a variety of cases. Two possible cases—graphical sag and display
case—are discussed as follows.
𝑋1 −∆𝑥
𝑌1 − ∆𝑦 = 𝑎 ∗ 𝑐𝑜𝑠ℎ ( 𝑎
)
𝑋 −∆𝑥
𝑌2 − ∆𝑦 = 𝑎 ∗ 𝑐𝑜𝑠ℎ ( 2 𝑎 ) (6)
𝑋 −∆𝑥
{𝑌3 − ∆𝑦 = 𝑎 ∗ 𝑐𝑜𝑠ℎ ( 3 𝑎 )
Graphical Sag. The ‘Graphical Sag’ functionality in PLS-CADD® requires known end points,
(X1, Y1), (X2, Y2), and a graphically selected middle point, (X3, Y3) (Figure 2). With three sets of known
coordinates, three equations, and three remaining unknowns: a, Δx, and Δy, there exists a unique solution
which can be solved for iteratively. The tension is then simply calculated using the catenary constant, a.
Note that for the symmetric catenary: Y1 = Y2 and |X1 – X3| = |X2 – X3|. Knowing this, the equations simplify
nicely. In short, the graphical sag case requires known end points and a known third point.
Display Case. Conversely, when attempting to display a wire at a known tension or corresponding
catenary constant, a, and known end points, (X1, Y1), (X2, Y2), the remaining four unknowns: (X3, Y3), Δx,
and Δy can be solved using equations (5) and (6). The curve can then be drawn through a back-calculated
middle point, (X3, Y3). In short, the display case requires known end points and known tension.
The catenary constant, a, is a very useful value and is quite important to understand. The value
merely describes the inversely related terms of horizontal tension, FH (force), and cable unit weight, w
(force/length), and thus has corresponding units of length.
𝐹𝐻
𝑎= 𝑤
(7)
Figure 4 shows the catenary equation with increasing catenary constants equal to 1 through 10.
Notice that curvature decreases as the catenary constant increases. Hanging conductors in transmission lines
have relatively “flat” curvature, meaning a very high catenary constant—perhaps 300-3,000 m (~1,000-
10,000 ft). This means that the theoretical datum for a typical span lies thousands of feet below the surface
of the earth. Remember, the catenary constant should not be confused with distance of cable to ground. The
catenary constant is simply a mathematical tool denoting a local origin of coordinates directly below the
low point or “belly” of the curve at a distance equal to the catenary constant. Although challenging to
calculate by hand, it is very important to understand that computers first calculate the catenary constant in
order to display wires or print sag-tension values for a variety of temperature conditions. Thus, the catenary
constant is a critical intermediate value for determining sag-tension relationships.
Page 21 of 284
16
14
y = a cosh(x/a)
12
10
a = 10
a=9
8
y
a=8
a=7
6
a=6
a=5
4
a=4
a=3
2
a=2
© 2017 D. Gelder a=1
0
-8 -6 -4 -2 0 2 4 6 8
x
Many useful relationships may be derived from the equation of position including slope, which is
the derivative of position with respect to x:
𝑑𝑦 𝑠(𝑥)
𝑑𝑥
= 𝑠𝑖𝑛ℎ(𝑥/𝑎) = 𝑎
(8)
For the symmetric catenary, mid-span sag is formulated based on the equation of position:
0.5𝜆
ℎ𝑥=0 = 𝑦(0.5𝜆) − 𝑦(0) = 𝑎 ∗ [𝑐𝑜𝑠ℎ ( 𝑎
)− 1] (10)
Finally, the following tension equations based on the horizontal tension, FH, are useful (PLS-CADD 1993):
Figure 5 shows a catenary free-body diagram cut at the low point or “belly” with end reactions.
The figure shows an effective vertical force equal to the weight of the cable acting a distance d from end
point J. It is interesting to note that the force passes through the intersection of lines f1(x) and f2(x), defined
by the slope of the curve at point O and end point J, respectively. This is significant because the geometry
Page 22 of 284
of the curve confirms that summing the moments about point O is equal to zero, as originally postulated
regarding the catenary (i.e. tension only, zero bending stiffness).
FV
FH
J
h w
FH
O
d
λ2
The formulation of the equations defining elastic behavior, including linear stretch and linear
thermal expansion, are presented herein. The elastic catenary equations are of extreme importance in
transmission engineering. Transmission structures are designed to support a variety of loading
combinations for the wire—including temperature, wind, and ice—to maintain proper strength and
clearances “at all times” (NESC 2012). To simplify the discussion, the effects of ice and wind are not
considered herein. It is only noted that ice increases the unit weight in the same plane as gravity, while wind
increases the unit weight out-of-plane. Thus, an effective unit weight may be calculated as the resultant of
self-weight, ice, and wind, deforming still in the shape of a catenary, but swung out-of-plane.
One approach for calculating the sag-tension catenary relationship is presented as follows. The
formulation of equations includes equations of compatibility, geometry, conservation of mass, and
equilibrium.
Page 23 of 284
Compatibility. The following equation utilizes the concept of an “unstressed” cable length,
LUnstressed, in which there is zero tension and at a reference temperature. The cable length, L, can then be
calculated based on changes in tension, ΔLTension, and/or temperature, ΔLTemp.
The following equations for linear force-displacement are based on Hooke’s Law where k is cable
stiffness, E is modulus of elasticity, and A is cross-sectional area. Note that positive force implies tension
increase and negative force implies tension decrease.
∆𝐿 𝑇𝑒𝑛𝑠𝑖𝑜𝑛 ≈ 𝐹𝐻 /𝑘 (14)
𝑘 = 𝐸𝐴/𝐿𝑈𝑛𝑠𝑡𝑟𝑒𝑠𝑠𝑒𝑑 (15)
The following equation is for linear thermal expansion where α is the coefficient of thermal
expansion, TEMP is the conductor temperature, and TEMPREF is the reference conductor temperature
(typically 0°C or 32°F) (SAPS 1997).
𝐻 𝐹
𝐿 = 𝐿𝑈𝑛𝑠𝑡𝑟𝑒𝑠𝑠𝑒𝑑 ∗ [1 + 𝐸𝐴 + 𝛼 ∗ (𝑇𝐸𝑀𝑃 − 𝑇𝐸𝑀𝑃𝑅𝐸𝐹 )] (17)
Geometry. The conductor length, L, of a symmetric catenary may be defined using the formula for
arc length, where x is equal to half of the span length, λ, and multiplying the arc length by two:
0.5𝜆
𝐿 = 2 ∗ 𝑎 ∗ 𝑠𝑖𝑛ℎ ( 𝑎
) (18)
Conservation of Mass. Based on the law of conservation of mass the total cable weight with fixed
ends, W, is constant regardless of changes due to tension or temperature. Thus,
𝑊 =𝐿∗𝑤 (19)
Solution. These equations can be combined into a solvable system of three equations with three
unknowns: L, FH, and a. These variables may be solved for iteratively using a spreadsheet or another tool.
𝐻 𝐹
𝐿 = 𝐿𝑈𝑛𝑠𝑡𝑟𝑒𝑠𝑠𝑒𝑑 ∗ [1 + 𝐸𝐴 + 𝛼 ∗ (𝑇𝐸𝑀𝑃 − 𝑇𝐸𝑀𝑃𝑅𝐸𝐹 )]
0.5𝜆
𝐿 = 2 ∗ 𝑎 ∗ 𝑠𝑖𝑛ℎ ( ) (20)
𝑎
𝑊
{ 𝐿= 𝑤
Equilibrium. The solution may be checked using an observation that the effective moment in
balance, M, equal to the product of horizontal tension and sag, is approximately constant regardless of
Page 24 of 284
changes in tension or temperature. This is essentially the same conclusion as proposed by the general cable
theorem which states that, “any point multiplied by the cable sag at this point is equal in magnitude to the
moment at the corresponding position in a simply supported beam of the same length and carrying the same
vertical loads” (Muvdi et al. 1997).
𝑀 = 𝐹𝐻 ∗ ℎ ≈ 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 (21)
To demonstrate that these equations may be used in a spreadsheet to obtain valid results, sample
sag-tension data is provided. The data is for a 122 m (400 ft) span of Marigold AAC conductor strung at
8,900 N (2,000 lbs) at 15.6°C (60°F). Input data is listed in Table 1. Sag-tension results for a variety of
temperatures are listed in Table 2. Results from PLS-CADD® are also listed in Table 2 for comparison.
Comparing the last four columns in Table 2 for horizontal force, FH, and sag, h, shows that results
corroborate.
Page 25 of 284
Table 2: Sag-Tension Data Produced with Spreadsheet for Marigold AAC Conductor
TEMP ΔL_Tension ΔL_Temp L a w FH h FH (PLS) h (PLS)
°C (°F) m (ft) m (ft) m (ft) m (ft) N/m (lbs/ft) kN (kips) m (ft) kN (kips) m (ft)
-51 (-60) 0.06 (0.20) -0.14 (-0.47) 121.98 (400.2) 1093 (3586) 15.27 (1.046) 16.7 (3.8) 1.7 (5.6) 16.8 (3.8) 1.7 (5.6)
-46 (-50) 0.06 (0.18) -0.13 (-0.42) 121.99 (400.2) 1009 (3311) 15.27 (1.046) 15.4 (3.5) 1.8 (6.0) 15.5 (3.5) 1.8 (6.0)
-40 (-40) 0.05 (0.17) -0.11 (-0.37) 122.01 (400.3) 939 (3081) 15.27 (1.046) 14.3 (3.2) 2.0 (6.5) 14.4 (3.2) 2.0 (6.5)
-34 (-30) 0.05 (0.16) -0.10 (-0.32) 122.02 (400.3) 877 (2878) 15.27 (1.046) 13.4 (3.0) 2.1 (7.0) 13.4 (3.0) 2.1 (6.9)
-29 (-20) 0.05 (0.15) -0.08 (-0.27) 122.03 (400.4) 824 (2704) 15.26 (1.046) 12.6 (2.8) 2.3 (7.4) 12.6 (2.8) 2.2 (7.4)
-23 (-10) 0.04 (0.14) -0.07 (-0.22) 122.04 (400.4) 783 (2567) 15.26 (1.046) 11.9 (2.7) 2.4 (7.8) 11.9 (2.7) 2.4 (7.8)
-18 (0) 0.04 (0.14) -0.05 (-0.16) 122.06 (400.4) 743 (2438) 15.26 (1.046) 11.3 (2.5) 2.5 (8.2) 11.3 (2.5) 2.5 (8.2)
-12 (10) 0.04 (0.13) -0.03 (-0.11) 122.07 (400.5) 706 (2317) 15.26 (1.046) 10.8 (2.4) 2.6 (8.6) 10.8 (2.4) 2.6 (8.6)
-7 (20) 0.04 (0.12) -0.02 (-0.06) 122.08 (400.5) 676 (2218) 15.26 (1.045) 10.3 (2.3) 2.8 (9.0) 10.3 (2.3) 2.7 (9.0)
-1 (30) 0.04 (0.12) -0.00 (-0.01) 122.10 (400.6) 649 (2129) 15.26 (1.045) 9.9 (2.2) 2.9 (9.4) 9.9 (2.2) 2.9 (9.4)
4 (40) 0.03 (0.11) 0.01 (0.04) 122.11 (400.6) 625 (2049) 15.25 (1.045) 9.5 (2.1) 3.0 (9.8) 9.6 (2.1) 3.0 (9.7)
Page 26 of 284
10 (50) 0.03 (0.11) 0.03 (0.09) 122.13 (400.7) 603 (1978) 15.25 (1.045) 9.2 (2.1) 3.1 (10.1) 9.2 (2.1) 3.1 (10.1)
16 (60) 0.03 (0.11) 0.04 (0.14) 122.14 (400.7) 583 (1912) 15.25 (1.045) 8.9 (2.0) 3.2 (10.5) 8.9 (2.0) 3.2 (10.5)
21 (70) 0.03 (0.10) 0.06 (0.19) 122.16 (400.8) 565 (1853) 15.25 (1.045) 8.6 (1.9) 3.3 (10.8) 8.6 (1.9) 3.3 (10.8)
27 (80) 0.03 (0.10) 0.07 (0.25) 122.17 (400.8) 548 (1798) 15.25 (1.045) 8.4 (1.9) 3.4 (11.1) 8.4 (1.9) 3.4 (11.1)
32 (90) 0.03 (0.10) 0.09 (0.30) 122.19 (400.9) 534 (1751) 15.24 (1.045) 8.1 (1.8) 3.5 (11.4) 8.1 (1.8) 3.5 (11.4)
38 (100) 0.03 (0.09) 0.11 (0.35) 122.20 (400.9) 519 (1703) 15.24 (1.044) 7.9 (1.8) 3.6 (11.8) 7.9 (1.8) 3.6 (11.8)
43 (110) 0.03 (0.09) 0.12 (0.40) 122.22 (401.0) 505 (1658) 15.24 (1.044) 7.7 (1.7) 3.7 (12.1) 7.7 (1.7) 3.7 (12.1)
49 (120) 0.03 (0.09) 0.14 (0.45) 122.23 (401.0) 493 (1617) 15.24 (1.044) 7.5 (1.7) 3.8 (12.4) 7.5 (1.7) 3.8 (12.4)
54 (130) 0.03 (0.09) 0.15 (0.50) 122.25 (401.1) 482 (1581) 15.24 (1.044) 7.3 (1.7) 3.9 (12.7) 7.3 (1.7) 3.9 (12.7)
60 (140) 0.03 (0.09) 0.17 (0.55) 122.26 (401.1) 471 (1545) 15.24 (1.044) 7.2 (1.6) 4.0 (13.0) 7.2 (1.6) 3.9 (13.0)
66 (150) 0.03 (0.08) 0.18 (0.60) 122.28 (401.2) 461 (1512) 15.23 (1.044) 7.0 (1.6) 4.0 (13.2) 7.0 (1.6) 4.0 (13.2)
71 (160) 0.03 (0.08) 0.20 (0.66) 122.29 (401.2) 451 (1481) 15.23 (1.044) 6.9 (1.5) 4.1 (13.5) 6.9 (1.5) 4.1 (13.5)
77 (170) 0.02 (0.08) 0.22 (0.71) 122.31 (401.3) 442 (1452) 15.23 (1.044) 6.7 (1.5) 4.2 (13.8) 6.7 (1.5) 4.2 (13.8)
82 (180) 0.02 (0.08) 0.23 (0.76) 122.32 (401.3) 434 (1424) 15.23 (1.043) 6.6 (1.5) 4.3 (14.1) 6.6 (1.5) 4.3 (14.1)
88 (190) 0.02 (0.08) 0.25 (0.81) 122.34 (401.4) 426 (1398) 15.23 (1.043) 6.5 (1.5) 4.4 (14.3) 6.5 (1.5) 4.4 (14.3)
93 (200) 0.02 (0.08) 0.26 (0.86) 122.35 (401.4) 419 (1373) 15.22 (1.043) 6.4 (1.4) 4.4 (14.6) 6.4 (1.4) 4.4 (14.6)
NONLINEAR BEHAVIOR (ADVANCED MECHANICS)
The catenary further complicates itself when considering nonlinear effects including creep, material
nonlinearities, bimetallic or composite material, flexible end supports, etc. These effects are the substance
of advanced mechanics of materials and are of great importance in transmission engineering. Many
nonlinear effects exceed the typical scope of hand-calculations and require substantial programming to be
utilized effectively. Additionally, to capture nonlinear effects, numerical models—such as finite element—
are often required. Thus, the formulation of nonlinear equations is not within the scope of this paper. Rather,
the concepts and some results produced using PLS-CADD® software are briefly presented.
Of great importance is permanent cable deformation due to excessive load and/or creep. While
several approaches exist, PLS-CADD® uses fourth order polynomials to model both stress-strain and creep.
The stress-strain of bimetallic conductors is treated as a weighted-average based on cross-sectional area.
𝜎 = 𝑘0 + 𝑘1 𝜀 + 𝑘2 𝜀 2 + 𝑘3 𝜀 3 + 𝑘4 𝜀 4 (22)
Figure 6 shows horizontal tension vs. conductor sag relationship. Figure 7 shows conductor sag vs.
conductor temperature. Figure 8 shows horizontal tension vs. conductor temperature. Each figure plots
“Linear Material,” as well as “Nonlinear Material” curves for “Initial” and “Creep” conditions.
Axis Title
0 ft 3 ft 6 ft 9 ft 12 ft 15 ft 18 ft 21 ft
20 kN
Linear Material
Nonlinear Material, Initial Condition
Nonlinear Material, Creep Condition
4 kips
16 kN
3 kips
Horizontal Tension
12 kN
Axis Title
y = 28301x-0.998
R² = 1
2 kips
8 kN
Page 27 of 284
Axis Title
-150 °F -50 °F 50 °F 150 °F 250 °F 350 °F 450 °F
7m
Linear Material
Nonlinear Material, Initial Condition
Nonlinear Material, Creep Condition 21 ft
6m
18 ft
5m
15 ft
Conductor Sag
4m
Axis Title
12 ft
3m
9 ft
2m
6 ft
Conductor: Marigold AAC 1113 kcmil
Span: 122 m (400 ft)
1m Sag Temp.: 15.6 °C (60 °F) 3 ft
Sag Tension: 8.9 kN (2.0 kips)
Unit Weight: 15.25 N/m (1.045 lbs/ft)
© 2017 D. Gelder. All Rights Reserved. Ult. Tension: 87.6 kN (19.7 kips)
0m 0 ft
-100 °C -50 °C 0 °C 50 °C 100 °C 150 °C 200 °C 250 °C
Conductor Temperature
Axis Title
-150 °F -50 °F 50 °F 150 °F 250 °F 350 °F 450 °F
20 kN
Linear Material
Nonlinear Material, Initial Condition
Nonlinear Material, Creep Condition
4 kips
16 kN
3 kips
Horizontal Tension
12 kN
Axis Title
2 kips
8 kN
Page 28 of 284
In transmission engineering, the rigidity of end supports is also of great interest—particularly when
modeling existing transmission lines. Central to this discussion is the timeless lesson: boundary conditions
matter. Various models for wire-to-structure boundary conditions have been proposed for multi-span
problems including the classic Ruling Span method (Thayer 1924). It is important to note that Ruling Span
was originally intended for deriving stringing tensions, in which wires are pulled through roller
attachments. Once strung, wires are clipped to the insulators, thus engaging the structure. ‘Pulley’ and ‘clip’
elements have been proposed to capture the various effects of clipping wires to insulators and engaging
structure flexibility (Peyrot and Goulois 1978; Peyrot and Goulois 1979; McDonald and Peyrot 1990).
Today’s most commonly used transmission design software, PLS-CADD® (1993), provides a means of
analyzing transmission lines under four different sets of boundary condition assumptions. These “analysis
levels” each have assumptions which may produce very different results in terms of structure loads from
wires. Scenarios provoking differences include: short-span-long-span, heavy angles, steep terrain, true
dead-ending, false dead-ending, single-span, guying, taps/intersections, flexible or rigid attachments, etc.
In current practice, many engineers modeling transmission lines typically select a single analysis level to
be used for checking designs. This is sufficient for preliminary design, but may expose final results to
pitfalls of the selected analysis level. As an alternative, it is recommended that transmission engineers
compare and contrast results using multiple PLS-CADD® analysis levels and employ good engineering
judgment to control variation in results.
CONCLUSION
The goals of this paper were to present the catenary equations of position, tension, and elongation
in a manner useful to a transmission engineering audience and to make equations seem less intimidating to
transmission engineers. These equations are of great importance to the design and analysis of transmission
lines. As transmission engineers explore, understand, and use these equations, minds will be empowered
and designs will improve.
Page 29 of 284
REFERENCES
Hibbeler, R.C. (2007). Engineering Mechanics Statics, 11th Ed., Pearson Prentice Hall, New Jersey.
McDonald, B. and Peyrot, A. (1990), “Sag-Tension Calculations Valid for Any Line Geometry.” Journal
of Structural Engineering, 116(9), 2374-2386.
Muvdi, B., Al-Khafaji, A., and McNabb, J. (1997). Statics for Engineers, Springer-Verlag, New York, NY.
Peyrot, A.H and Goulois, A.M. (1978), “Analysis of Flexible Transmission Lines.” Journal of the Structural
Division, ASCE, 104(ST5), 763-779.
Peyrot, A.H. and Goulois, A.M. (1979), “Analysis of Cable Structures.” Computers and Structures, 10(5),
805-813.
NESC (2012 and older editions), National Electrical Safety Code, IEEE, New York, NY.
PLS-CADD (1993 and newer versions), A Computer Program for the Integrated Analysis and Design of a
Transmission Lines (including spotting optimization), Power Line Systems, Madison, WI.
SAPS (1997), A Computer Program for the Nonlinear Structural Analysis of Power and Communication
Systems, Power Line Systems, Madison, WI.
Thayer, E.S. (1924), “Computing tensions in transmission lines”, Electrical World, 84(2), 72-73.
Weisstein, E. W. (1999), “Catenary.” MathWorld—A Wolfram Web Resource,
<http://mathworld.wolfram.com/Catenary.html> (June 9, 2015).
NOMENCLATURE
Page 30 of 284
50th Annual
Transmission and Substation Design and
Operation Symposium
By
Page 31 of 284
Abstract
For decades, power factor / dissipation factor testing has been used throughout the electrical industry
and has triggered the attention of many who are directly involved in electrical equipment operation,
maintenance and asset management. The predominant question today is no longer about how to perform
this test but how to move forward with this technology understanding its advantages, limitations and the
way to overcome such limitations.
It has been over a century since the first insulation material was used to isolate two different potential
levels. Different methods have become available in the form of DC and AC techniques to quantify dielectric
losses within the insulation material. The power factor/dissipation factor testing technique has evolved
accompanied by the development of more sophisticated and reliable electronic components and IT
solutions. National and international standards now address power factor testing as a routine method for
insulation condition assessment and provide reference values establishing boundaries to guide operators
and maintenance personnel on when to resume operation, perform additional investigation, or remove from
operation. Trending analysis, however, remains the most reliable approach to analyze power factor results
but is still not without limitations. No matter the analytic approach used, a single power factor measurement
at line-frequency may mask important information needed to make sound technical recommendations and
avoid unwarranted actions that may compromise the reliability of the operation.
Line-frequency power factor values can increase due to moisture contamination, insulation degradation,
aging or mechanical damage, but it is quite complicated to determine the cause for elevated (or too low)
line-frequency power factor values. Moreover, power factor is temperature and frequency dependent;
therefore, variations of temperature and frequency may lead to incorrect interpretation of results. Proper
correction is needed. The use of the Individual Temperature Correction (ITC) algorithm is presented in this
paper giving field operators a realistic correction based on the unique dielectric response of the specimen
under test.
Linearity and non-linearity of the dielectric system may unknowingly change the analysis approach
from one only comprised of dielectric loses to another one including loses due to discharge activity. The
voltage dependence detection (VDD) algorithm is discussed as an evaluation tool in the power factor test
to identify non-linearity of the dielectric system.
In addition to power factor testing of the liquid-impregnated insulation, guidelines are provided for
liquid insulation dissipation factor test in the field and the importance of advanced applications of line-
frequency power factor test in the frequency domain spectroscopy analysis is reviewed.
This paper summarizes the latest development in power factor testing for insulation assessment of
power and distribution transformers in the field. Comprehensive guidelines described in the document will
improve the decision making process, will clarify doubtable results, will allow prioritization of maintenance
activities and will prevent failure and accelerated degradation of the insulation system.
Page 32 of 284
I. INTRODUCTION
An insulation system is represented in electric circuits by a capacitor. Two electrodes separated by a
dielectric form a capacitor same as two conductive surfaces separated by an insulating material form an
insulating system. A perfect or ideal capacitor would have no other component of current except the
capacitance component. However, no dielectric materials, not even vacuums, are perfect. There are always
nonconforming electrons that form part of the conduction or loss component of current. The number of free
electrons depends on the molecular structure of the material, contaminants (particularly moisture), and
temperature.
For power transformers, the service life is estimated based on the analysis of the capacitive components
or the dielectric insulation system and dielectric mode failures are the highest contributor of failure modes,
irrespective of transformer application. Dielectric mode failures on fluid-immersed transformers may result
in fire, explosions and damage to surrounding equipment and/or personnel.
To assess the condition of the dielectric system, off-line dielectric response techniques are readily
available and are commonly used. The dielectric response can be obtained in the time domain (DC
Insulation resistance, polarization Index, polarization-depolarization current) and/or in the frequency
domain (line frequency power factor and frequency domain spectroscopy).
Historically, the DC insulation resistance method pioneered the industry of insulation diagnostics and
has evolved for over a century. The portability of DC equipment made it ideal for field condition assessment
of medium, high and extra-high voltage apparatuses. AC testing equipment were bulky and difficult to
transport, but the continuous development of power electronics, communications and materials technology
has promoted the AC testing equipment forward and its application for insulation diagnostics has spread
globally.
The electrical power industry has relied on single frequency (line frequency) power factor/ dissipation
factor and capacitance testing to assess electrical insulation for almost a century. Industry specialists,
researchers and instrument manufacturers have worked together to provide advanced features to improve
and expand the information obtained from line-frequency power factor testing on fluid-filled transformers.
Insulation resistance tests are usually conducted at constant direct voltages of 100 – 15,000 V having
negative polarity and measurements can go from the megaohm (M) range up to teraohms (T). Negative
polarity is preferred to accommodate the phenomenon of electro-endosmosis [1]. Readings of insulation
resistance are taken after the test direct voltage has been applied for 1 minute. Most important to understand
is that for insulation resistance (IR), there is no need to overstress the insulation material, the idea is to
Page 33 of 284
measure the dielectric properties of the capacitance under test. The amount of power produced is very low,
the test is considered non-destructive as well as non-intrusive.
IR test is recommended during commissioning and after protection trip due to internal fault in power
and distributions transformers [2]. IR tests are made to determine the insulation resistance magnitude from
individual windings to ground or between individual windings. In transformers, the recommended practice
in measuring IR is to always ground the tank (and the core). Short-circuit each winding of the transformer
at the bushing terminals. IR measurements are then made between each winding and all other windings
grounded.
a. Two-winding transformer
o (HV + LV) to GND
o HV to (LV + GND)
o LV to (HV + GND)
b. Three-winding transformer
o HV to (LV + TV + GND)
o LV to (HV + TV + GND)
o (HV + LV + TV) to GND
o TV to (HV + LV + GND)
Notice that on the recommended hook-up connections above, “guard” is not used. The guard line is at
the same potential as the DC source generator potential. For a HV winding, 5kVDC might be OK, but not
for a 480V LV winding.
Interpretation of IR results is based on historical data, transformer design, type of insulating material,
and contamination level. Transformers filled with natural ester-based liquid generally have lower insulation
resistance compared with ones filled with mineral oil [2]. Trending analysis of IR results is more significant
than the instantaneous measurement; keep a record of the voltage used for testing and the obtained values
corrected to 20°C.
Figure 1 shows two different specimens tested for IR. One has a significant change of IR from high G
values while the other one has not changed significantly and numbers remain steady even in the M range.
The insulation temperature also affects the readings, typically testing at higher temperature results in lower
insulation resistance.
Page 34 of 284
Figure 1 Trending of Insulation Resistance values of two different specimens.
Very low values of IR are indicative of a breakdown in the insulation to ground or between windings,
high moisture, or heavy carbon tracking. Additional tests such as polarization index, insulation PF, or
dielectric frequency response are required to validate and better determine the reason for low IR values.
Page 35 of 284
Dielectric loss is the energy lost or released as heat when an electrostatic field is present across an
insulation system. Losses can be broadly classified as conductive losses, arising from leakage current, or
polarization losses. The total losses (IR ∙ V) in a dielectric are equal to the combined polarization losses and
conductive losses present.
Dielectrics perform best when they are clean, dry, relatively void-free, and utilized within a certain
temperature range. Adversaries to a dielectric’s continued good health are heat, moisture and oxygen.
Continuous degradation of the insulation system is observed when power factor/dissipation factor test
reports higher dielectric loses. Normal transformer service shows a slow increase of these dielectric losses
and that is expected to see during normal and/or routine maintenance testing. Nevertheless, rapid increase
of losses are indication of an active failure condition that may lead into dielectric breakdown.
Power Factor (1) is the cosine of (the complementary angle of the “loss angle”) while Dissipation
Factor (2) is the tangent of , or the loss angle. As described in [2], the normal in-service and new PF limit
for mineral-oil-filled power transformers < 230 kV is 0.5% PF at 20 °C, and the normal and new limit for
transformers 230 kV is 0.4%. To help reduce the risk of catastrophic failure, the limit for serviceability
of all mineral-oil-filled transformers is 1.0% PF at 20 °C. PFs between 0.5% and 1.0% at 20 °C require
additional testing and investigation to confirm that a problem is not worsening.
IR
Power Factor: % PF cos 100 % 100 % (1)
IT
IR
Dissipation Factor: % DF tg 100 % 100 % (2)
IC
Power factor/ dissipation factor represents the average condition of the total insulation system under
test. If the system is big enough, some localized contamination may be impossible to see. On a large single-
phase autotransformer with an embedded tertiary with no external access, only one power factor/dissipation
factor and capacitance measurement can be performed on such a transformer (all the winding terminals are
connected together and energized for a winding to ground insulation measurement). Degradation or
contamination might be difficult to pinpoint.
Increasing levels of some contaminants might not provide notable change in power factor/ dissipation
factor at line frequency and depending on the conductivity of the material, the effect of contamination may
be observed at a frequency different from 50/60Hz.
To overcome this limitation, a dielectric frequency sweep in the range between 1 – 500 Hz is required.
Page 36 of 284
B. Thermal dependence
On this point and from IEEE Std. C57.12.90, the following is extracted as published:
“NOTE 3— b) Experience has shown that the variation in power factor with temperature is substantial
and erratic so that no single correction curve will fit all cases. The power factor shall be reported along with
the top oil temperature measured and the bottom oil temperature if available. No temperature correction
shall be applied. Temperature correction of the power factor results for trending basis may be applied by
the purchaser”.
“…The loss factor of insulation can be sensitive to variations in temperature, in which case a correction
factor needs to be applied to measured values. This is done to allow comparison of tests performed at
different temperatures. The reference temperature used is 20 °C. Correction factors may be available from
equipment manufacturers and test equipment manufacturers. (Additional correction factors were available
in IEEE Std. C57.12.90-2006, 10.10.5, which was removed in IEEE Std. C57.12.90-2010.)...”
The standards clearly indicate that measurements of losses in the dielectric material are sensitive to
temperature variation as well as dielectric condition changes. Changes in power factor due to moisture,
contamination or degradation of the insulation material is unique to the asset and cannot be duplicated in
another asset, even one with an identical nameplate, because of different insulation condition, electrical,
thermal and/or mechanical stress, operation condition, maintenance work, etc.
To overcome this limitation, an algorithm to estimate the individual temperature correction is required.
C. Voltage Dependence
The voltage dependence phenomenon in solid insulation during line-frequency PF/DF test is well
described in [3] and [4]. The losses measured in the insulation system when a line-frequency (50/60Hz)
signal is applied, is a composite of dielectric losses which are constant with voltage and power loss due to
discharges. Mathematically this can be expressed as the total conductance of the system in (3)
1
f
2 f
G C 0 tan 0
Vm
Q i sin t i (3)
0
Based on (3), if no discharges occur, the amount of charge increase on the electrodes or conductors as
a result of internal discharges (Qi) equals zero. To visualize this effect, a narrow band DFR test at different
voltage levels carried out on epoxy type MV equipment is presented in Figure 3.
Page 37 of 284
Figure 3 DFR obtained from MV epoxy-type insulation specimen at different voltages
It is clear from Figure 3 that the voltage dependence of an insulation system can not be determined by
a single voltage test. In the specific case of oil-impregnated insulation, the expected condition is to measure
PF/DF at line frequency at a single voltage and that the value under ideal conditions will not change. A
deviation from this statement will indicate that the ideal condition has been altered and potential degradation
is taking place within the insulation system of the transformer. Figure 4 shows an ideal condition of the
insulation inside the transformer in a tip-up test.
Figure 4 Tip up test carried out on a 1991 Dyn1, 67/13.09 kV, 25MVA
The limitation of power factor test at line frequency is to have a single voltage test value for a specific
capacitive zone. To overcome this limitation, a technical approach to determine voltage dependency is
required. %VDF (percentage voltage dependence factor) is a very small value, typically below 0.1% for
linear %PF response.
Page 38 of 284
V. DEVELOPMENT OF INSULATION DIAGNOSTICS ON POWER TRANSFORMERS
As discussed in section IV.A., the single frequency power factor measurement has limited information
regarding the factor affecting the insulation system. Moreover, similar PF values might be obtained out of
different specimens even though the insulation condition is different.
Figure 5 describes six different examples of insulation conditions where line frequency power factor is
the same or very close on all of them. The reality shows that a dielectric frequency response between 1 and
500 Hz helps visualization of potential factors affecting the condition of the insulation.
The research presented in [5] shows the importance of the narrow band dielectric response. The specific
application on oil-impregnated paper (OIP) HV bushings allows a better assessment of the insulation at
1Hz. It is recommended to plot the narrow band dielectric response in a logarithmic scale (as shown in
Figure 5), otherwise the smallest decade (1 – 10Hz) is not clearly visualized.
Narrow band DFR from 1 to 500Hz has been already implemented in state-of-the-art power factor test
sets and its use has more applications as it is discussed in the next section.
As mentioned in the previous section, narrow band DFR has an additional application when used for
advanced diagnostics in power factor testing.
Figure 6 clearly shows how the dielectric response in the frequency domain and in the thermal domain
change based on the condition of the insulation system and therefore the use of tables and average correction
factor values might be misleading and not accurate.
Page 39 of 284
(a) (b)
Figure 6 Influence of temperature on: (a) DFR of insulation with 1% mc; (b) DFR of insulation with 3.5% mc
The obtained narrow band DFR from the inter-winding insulation is used now to determine the
individual temperature correction of the line frequency power factor measured at temperatures different
from 20°C.
The mathematical approach used to correlate line-frequency power factor with temperature is an
Arrhenius-based equation (4). This approach describes the ‘frequency shift’ factor, which is dependent on
the temperature difference between the temperature of the measurement T2 and the reference temperature
T1 (expressed in Kelvin).
Ex,y 1 1
T (4)
A x , y T1 , T 2 e
kB 2 T 1
The equation considers an exponential function related to temperature (T), Boltzmann constant (kB)
and the activation energy value (Ex,y) of the material. It has been found that the shape of the dielectric
response (PF/DF versus frequency) does not change very drastically with temperature for quite a large
group of solid dielectric materials; rather, as temperature changes, the response (a spectral shape) shifts
with respect to frequency while remaining intact. This means that a PF/DF value measured at line frequency
and the temperature at time of test, is the exact same PF/DF that would be measured at the reference
temperature of 20°C with the big exception that it would occur at a frequency shifted from and different
than line frequency
Therefore, using narrow band DFR, the individual temperature correction (ITC) factor can be estimated
for the line frequency power factor based on the real insulation condition [6].
Page 40 of 284
C. Voltage Dependence Detection (VDD)
The effect of voltage over an insulation system is not observed by a single voltage, single frequency
power factor test. The voltage dependency effect has been observed and studied and the new power factor
technology uses the so called percentage voltage dependence factor (%VDF) representing the distortion of
the measured current with respect to its fundamental. The distortion in the measured current signal, as
represented in Figure 7, implies that the insulation system has lost its linearity [7].
Figure 7 Applied Voltage and measured current through a voltage dependent insulation system
Therefore, the %VDF can be calculated in a similar way as the total harmonic distortion (THD) of the
measured current signal as presented in (5):
2
Ih
h2
(5)
% VDF 100 %
I1
The application of the VDF is easily demonstrated testing solid insulation specimens such as epoxy
filled MV instrument transformers. It is clear the effect of voltage dependency and non-linearity of the
dielectric system by performing a tip-up test on the insulation and monitoring the %VDF as presented in
Figure 8. The ideal oil-paper insulation presented in Figure 4 compared to the plot in Figure 8 clearly
support the need to have implemented voltage dependence detection algorithms in the power factor testing
instrument.
Figure 8 (b) shows that as soon as the voltage reaches the point where the %PF loses linearity, the
%VDF follows the non-linear effect. If the insulation shows deterioration, PD activity or contamination,
%VDF provides a warning flag to the end user “suggesting” to add a tip-up test to the series of tests carried
out on that specific transformer.
Page 41 of 284
(a) (b)
Figure 8 Tip-up test on MV PT: (a) test hook-up; (b) %PF and %VDF as a function of test voltage
In the frequency domain, the dielectric spectrum is obtained applying a low voltage AC excitation
signal (140 Vrms) to the insulation system. The frequency spectrum is typically obtained between 1 kHz and
1 mHz as shown in Figure 9.
The response is analyzed based on mathematical modeling and comparative analysis against a well-
documented materials’ database to determine primarily:
In addition, DFR provides information of the presence of contaminants by non-typical responses and
the information can be converted from the frequency domain into the thermal domain providing the
dielectric thermal response of the insulation system as a plot of %PF or %DF vs. insulation temperature as
shown in Figure 10.
Page 42 of 284
Figure 10 Example of dielectric thermal response of a fluid-filled power transformer
The IEEE transformers committee is finalizing the work to generate the guidelines for DFR application
on power and distribution transformers. The document is expected to be published in 2018 under the
designation IEEE Std C57.161.
VII. SUMMARY
As presented in this article, condition assessment of transformer insulation is not a simple task. Limited
information obtained by a single frequency power factor may not be conclusive in the analysis and the
features developed are tools that need to be properly understood and applied.
Page 43 of 284
VIII. REFERENCES
[1] IEEE Recommended Practice for testing Insulation resistance of Rotating Machinery. IEEE Std
43-2000 (R2006).
[2] IEEE Guide for Diagnostic Field Testing of Fluid-Filled Power Transformers, Regulators, and
Reactors. IEEE Std C57.152TM-2013.
[3] IEEE Std. 286-2000 (R2012). “IEEE Recommended practice for measurement of power factor tip-
up of Electric machinery Stator Coil insulation”.
[4] T. W. Dakin, “The Relation of Capacitance Increase with High Voltages to Internal Electric
Discharges and Discharging Void Volume”. AIEE Transactions. Part III, Issue 3, 1959
[5] D. Robalino, I. Guener, P. Werelius, “Analysis of HV Bushing Insulation by Dielectric Frequency
Response”. Proceedings from the IEEE EIC Conference 2016.
[6] D. Robalino, “Individual Temperature Compensation – benefits of dielectric response
measurements”. Transformers magazine, Vol. 2, Issue 3. 2015.
[7] Dielectric Frequency Response Measurement Technology for Measurements in High Interference
AC and HVDC Substations”. 2011 TechCon Asia-Pacific, Sydney, Australia
Page 44 of 284
50th Annual
Transmission and Substation Design and
Operation Symposium
By
Page 45 of 284
Abstract
A transformer is one of the most important assets in a high voltage substation. The life
expectancy of a power transformer can be drastically reduced by the presence of internal moisture.
Excessive amounts of moisture in paper insulation can degrade the thermal characteristics and loading
capabilities of a power transformer. With the transformer aging, loading, thermal and electrical stresses,
degradation of solid insulation and other factors, the moisture content increases over time reducing the
dielectric strength of the insulation system.
Oil impregnated paper insulation is hygroscopic in nature and retains the maximum amount of
moisture when compared against the insulating oil. Depending on equipment availability and company
operating procedures, each utility uses different field dry out techniques such as hot air or oil
circulation, vacuum only or heat and vacuum methods to extract the moisture from the windings. The
dry out process helps in slowing down the rate of aging of the cellulose material. Low moisture levels as
recommended in different IEEE standards provide power system operators confidence in the dielectric
integrity of the transformer and its reliable operation.
Utility operators utilizing field dry out techniques often run into a challenge to determine the
remaining moisture and appropriate time to stop the process. There are traditional methods available to
estimate the moisture in solid insulation, however many of them rely upon a state of thermal
equilibrium or other charts; that only provide approximation of the remaining moisture in the insulation.
This uncertainty in moisture estimation could prompt operators to longer dry out durations and at times
lead to the risk of excessive dryness making paper insulation too brittle that reduces the tensile strength
and Degree of Polymerization.
This paper will introduce the Dielectric Frequency Response (DFR) measurement technique that
allows a user to determine the moisture content in solid insulation before, during and at the completion
of dry out process. It helps in determining efficiency of the process, indication of acceptable moisture
levels and appropriate time to stop the process. This paper will include field measurements on Extra
High Voltage (EHV) transformer and reactors with assessment of dry out results utilizing the DFR test
method. A comparative analysis of DFR technique against the traditional moisture estimation methods
will be provided to highlight the application of DFR during the dry out operation.
I. Introduction
Power Transformers reliable operation is very critical for the electric power system. The life of a
transformer is primarily based upon the health of its insulation system. Transformers are subjected to
electrical and thermal stresses throughout their life cycle. It is a well-known fact that moisture, oxygen
and temperature are the major factors that can cause accelerated aging and would drastically reduce
the insulation life [1]. Temperature is dependent on the loading and ambient conditions during
transformer operation. Operating a transformer at full load conditions and longtime emergency
overloading cycles should be controlled to avoid reaching limits of hot spot temperature as any increase
in temperature above those limits for an extended period of time can seriously accelerate the aging
process of the insulation. Oxygen can get into the transformer through gasket leaks, exposure to
atmosphere, the air breathing conservator, leaky bladder, as a byproduct of insulation degradation and
several other external sources [2]. Oxidation of the insulating oil leads to acid and sludge formation that
Page 46 of 284
further degrades the paper and obstructs the natural flow of oil, thus restricting the cooling process [3].
Oil preservation system and oil inhibitors are used to prevent oxygen from entering into the system,
control the levels of oxygen and its reaction with the insulating oil [4]. Moisture in transformer, even
though not desired, cannot be completely avoided. Cellulose insulation is hygroscopic in nature and has
great affinity to absorb and retain moisture. Moisture in solid insulation reduces the dielectric strength,
restrict the maximum thermal loading capabilities, accelerates the aging process, and causes
degradation of cellulose, which in turn, will produce more water as a byproduct of paper disintegration.
This leads to lower tensile strength and the insulation’s ability to handle mechanical stresses is
compromised. The oxygen moisture combination at high temperatures cause accelerated degradation of
the cellulose and directly impacts the remaining useful life of a transformer.
Although, aging of the insulation cannot be stopped, it can certainly be slowed down and
prolonged to extend the life of a transformer. Moisture is one of the factors that can be controlled. A
new transformer coming out of the factory is expected to have a moisture concentration at or below
0.5% of the weight of the solid insulation. Before commissioning, utility operators put all the efforts to
extract as much moisture as possible from the solid insulation to have a “dry” transformer. Oil filtration
techniques can take moisture and other contaminants out of the oil only. In order for moisture to be
extracted from solid insulation, usually a combination of high temperature and vacuum is used.
Different field dry out techniques such as hot air or oil circulation, vacuum only or heat and vacuum
methods are used to extract the moisture from the windings. Some methods use cold trap for moisture
extraction and to estimate the rate and volume of moisture extraction from the insulation.
One of the challenges experienced by utility operators is to correctly estimate the moisture
present in the solid insulation of a new or service aged transformer before and after the field dry out
process. An additional challenge encountered is to determine the efficiency of the dry out system and
optimize it so that the correct time to stop the process can be identified. Based upon the voltage class
rating of a transformer, guidelines are provided in IEEE C57.93 “Guide for Installation and Maintenance
of Liquid-Immersed Power Transformers” [5] for the durations related to vacuum hold time, minimum
set and energization time. However, since each transformer is unique and the initial moisture
concentration may be different, which along with efficiency of the dry out system would determine the
duration of the whole process to reach the desired final moisture concentration in solid insulation. New
and service aged transformers of same voltage class need to be treated differently as the amount of
moisture that can be extracted would differ. The recommendations made in the IEEE C57.93 standard
are general guidelines and are to be used in absence of any other recommendations set by a utility or a
manufacturer.
It is imperative to know the amount of moisture present before the dry out process is initiated.
Additionally, determining the moisture throughout the process is of great benefit. The traditional
methods used are tests relying upon a state of equilibrium or charts correlating moisture, vapor
pressure and temperature. The DFR technique for field applications has been in existence since early
1990’s and operators are now utilizing it to better estimate the moisture concentration during the dry
out operation. It not only helps in optimizing the process but also allows to prevent the risk of excessive
dryness making paper insulation too brittle and reducing its tensile strength.
Page 47 of 284
II. Dielectric Frequency Response
AC insulation Power Factor (PF) is a routine test performed on power and distribution
transformers to assess the condition of the insulation system. High PF values may point towards
presence of high moisture, contamination and possible deterioration of insulation as it ages. However, it
has been found that PF values do not directly relate to moisture present in the insulation system.
As discussed earlier, cellulose or paper has great affinity towards moisture or water. Per IEEE
C57.152-2013 “IEEE Guide for Diagnostic Field Testing of Fluid-Filled Power Transformers, Regulators,
and Reactors” water migrates between the solid and liquid insulation in a transformer with changes in
load and, therefore, temperature. Consequently, the concentration of water in the insulating liquid
alone, expressed in ppm, does not provide sufficient information to obtain an adequate evaluation of
the insulation system dryness [6]. There are moisture equilibrium curves available to estimate the
moisture in paper based upon the moisture in oil of an oil-paper based insulation system. IEEE Std
C57.106 “IEEE Guide for Acceptance and Maintenance of Insulating Mineral Oil in Electrical Equipment”
discusses the absence of thermodynamic equilibrium in an operating transformer. This leads to the
conclusion that a quantitative correlation of moisture in liquid to moisture in paper is not currently
feasible [7]. Cautions have been issued to not use these curves to determine the dryness level of the
solid insulation in an operating transformer [6]. Moisture concentration in paper is one of the critical
information that power system operators should know to maintain and extend the life of a transformer
insulation system.
Dielectric Frequency Response (DFR) also known as Frequency Domain Spectroscopy (FDS) is an
insulation diagnostic technique that uses the concept of insulation PF and capacitance measurement at
different frequencies (1 mHz -1000 Hz) to determine the dielectric response of the insulation [8]. The
results are used to determine the moisture content in cellulose (oil impregnated paper/pressboard)
insulation of power and distribution transformers.
A DFR measurement is typically performed in the field as per Figure 1. A step by step method to
perform DFR measurement is described in [9]. It measures the impedance of the system and with
impedance’s complex capacitance model, it calculates the capacitance, loss factor and PF at discrete
frequencies as shown in equation (1) to (5) to obtain a DFR curve between PF and frequency.
Page 48 of 284
1
𝑍= (1)
𝑗𝜔𝐶
𝐶 = 𝐶 ′ − 𝑗𝐶" (2)
1
𝐶′ = 𝑅𝑒 { } (3)
𝑗𝜔𝑍
1
𝐶" = −𝐼𝑚 { } (4)
𝑗𝜔𝑍
𝑅𝑒{𝑍}
𝑃𝐹 = (5)
|𝑍|
Where,
Z = Impedance
ω = Angular Frequency
C’ = Capacitance
C” = Loss Factor
PF = Power Factor
Power Factor
Frequency
Figure 2: DFR response of a two winding transformer with 0.4% moisture at 22 ⁰C
The DFR curve as shown in Figure 2 shows the response of a two winding transformer CHL
insulation with a moisture concentration of 0.4% at insulation temperature of 22 ⁰C. With its wide
frequency range, DFR has the advantage to separate the dielectric response of solid and liquid
insulation. The higher and lower frequencies response is primarily influenced by moisture in paper and
the middle frequencies predominantly show the dielectric loss response of the insulating liquid. The DFR
measurement is analyzed utilizing the modeling technique to provide percent moisture concentration
Page 49 of 284
present in the solid insulation along with capacitance, PF value at 60 Hz normalized to 20 ⁰C and oil
conductivity as shown in Figure 3.
The PF values are temperature dependent and since the DFR curve is basically PF measurements
at different frequencies, the DFR response is also temperature dependent. Temperature is a very
important parameter to be recorded for the DFR measurement. It helps in accurate estimation of bulk
moisture concentration present in the solid insulation. Additionally, the correct insulation temperature
recording allows selection of appropriate frequency range required to perform the test and obtain
reliable measurements.
DFR can be utilized as a powerful tool before, during and after the dry out operation to
determine the moisture present and increase the efficiency of the process. At the completion of final oil
fill and after a wait time of 24 hours to allow time for oil impregnation, DFR is recommended to be
performed on final oil-paper insulation system assembly. That measurement would provide operators a
reference moisture value that could be used for trending and future benchmarking.
Field dry out is conducted on power transformers and reactors to extract as much moisture as
possible absorbed in the oil impregnated paper due to atmospheric exposure, and during the
construction and assembly process. A combination of heat and vacuum cycle as shown in Figure 4 yields
the best results for moisture extraction. Vacuum pressure and inside temperature is constantly
monitored to determine the remaining moisture concentration in the paper insulation.
Page 50 of 284
Figure 4: A typical setup for field dry out system [11]
The following steps based upon reference [10] are conducted for the field dry out:
Insulating oil to be used during dry out is tested for moisture, specific gravity, interfacial tension,
dielectric strength and Dissolved Gas Analysis.
Winding temperature is measured by measuring the winding resistance (with 1 A at 50 V DC) on
any one phase before the start and throughout the dry out process. Reference temperature and
reference winding resistance reading should be known to calculate the winding temperature.
Insulating oil is preprocessed in the vacuum oil processor unit to remove excess water and gas.
It is passed through power heaters to heat up the oil to 60-65 ⁰C.
Vacuum pump is used to draw preliminary vacuum on the tank of the transformer
An auxiliary vacuum pump is used along with vacuum oil processor unit to reach acceptable
stable vacuum levels and maintain them during the final vacuum and final oil filling stages.
Vacuum degassed hot oil (50-60 ⁰C) is filled in the transformer tank to about 30-60% of
nameplate oil volume quantity.
The hot oil is circulated through the processing unit, discharging oil into the transformer at 60 ⁰C
from the top valve until the oil coming out of the transformer from the bottom drain valve has
reached 50 ⁰C.
The heat cycle is continued for six hours once 50 ⁰C mark is reached. This is done for the
preliminary vacuum to reach the required vacuum level and to reduce overall time duration for
processing.
One or more heat cycle may be required based upon the transformer dew point, length of
storage under dry air, or if the temperature is below the minimum requirement.
Oil is drained from the transformer to begin the first stage of high vacuum.
Page 51 of 284
Based upon piper chart as shown in Figure 5, vacuum is drawn to a suitable micron level
appropriate for the winding temperature.
Winding temperature and vacuum levels are recorded every 30 mins in the dry out log.
Stable vacuum is reached when the vacuum has not drifted more than 100 microns during a six-
hour period.
Perform a vacuum leak test as required or if vacuum levels have drifted more than 100 microns
in a six-hour period.
Continue vacuum set time for a period of time defined by the voltage class of the transformer.
Upon completion, using piper chart compare the stable vacuum level against the winding
temperature to estimate the moisture in cellulose insulation.
Final oil fill is done without any stoppage, interruption or break of the vacuum. Filling is done
through top valve and completed by closely monitoring the fill rate as to not allow transformer tank
vacuum to change more than 100 microns from the target micron level during filling. Final fill should
cease at 4-6 inches from the top of a conservator unit. The fill is allowed to go slightly over expected
volume to allow for oil contraction with cooling and when vacuum is relieved. At the end of final fill,
vacuum pull on the main tank is stopped by closing off the vacuum hose. Once the vacuum is broken, fill
of the conservator unit can be completed to finish the topping off process [10]. At the end, as per IEEE
C57.93 standard, based upon the voltage class of the transformer, a minimum amount of set time is
required before the unit can be energized. This time allows for full oil impregnation of the cellulose
insulation. The set time can also be used for electrical tests to be performed on the unit.
Page 52 of 284
A single phase 765 kV 750 MVA auto transformer field dry out results are shown below to
indicate the rate of change of temperature and vapor pressure during the vacuum cycle.
As shown in Figure 6, it can be noted that there was a change of more than 100 microns in a six-
hour period and in that condition the vacuum set time is reset back to 24 hours for 765 kV class
transformer. 24-hour vacuum set time was based upon utility guidelines. At the completion of vacuum
cycle, the vapor pressure value was 150 and temperature of the windings was close to 48 ⁰C. Based
upon the piper chart information, moisture in the paper insulation was estimated to be close to 0.20%.
Before beginning the dry out process, it is important to know the moisture currently present in
the paper insulation.
During the process, operators should be able to determine system efficiency by measuring:
o At what rate percentage moisture is coming down and
o How much moisture is remaining in the solid insulation.
Based upon those two factors, a decision can be made as when to stop the process.
Once the process is complete and final oil fill is finished, percentage moisture should be
measured. A stand time of 24 to 48 hrs should be allowed to measure the final moisture
concentration in solid insulation before the transformer is commissioned.
Page 53 of 284
DFR measurements can be used to achieve all the tasks mentioned above. The following section will
show case examples of application of DFR technique for process optimization. The DFR results are
compared with the existing methods to prove the validity and suitability of the method in the field
environment.
Dew point measurement was taken on a 765 kV, 100 MVAr single phase reactor before the dry out
process. Based upon the measurement of dew point, pressure and temperature inside the tank, the
average moisture at the surface of the insulation was estimated to be about 1.4 %.
Figure 7: DFR measurement before the dry out on a 765 kV single phase reactor
Page 54 of 284
DFR measurement was conducted before the dry out at 100 V with a frequency range of 1000
Hz to 2 mHz. Insulation temperature was 32 ⁰C. The DFR response is shown above in Figure 7. Based
upon the DFR curve and modeling analysis, the moisture in solid insulation was estimated to be 1.5 %.
The results compared very well with dew point measurement. One thing to note is that DFR
measurement doesn’t require any wait time for equilibrium as it provides the bulk moisture
concentration whereas for dew point method that provides the average surface moisture of the
cellulose insulation, enough time needs to be provided for equilibrium and stabilization of moisture
before a measurement can be taken. If this is followed, dew point and DFR results will be very
comparable.
It was decided to perform a field dry out on the unit using a combination of hot oil circulation
and vacuum cycles. Hot oil was circulated to bring the winding temperature to above 60 ⁰C. During the
vacuum cycles DFR data was obtained at an interval of every six hours. During the DFR measurement, it
is recommended to close the vacuum valve to prevent any vacuum pull during the measurement.
Following graph shows the dielectric response in the first 24 hours of vacuum cycle.
Page 55 of 284
Figure 9: DFR measurements in the first 24 hours of vacuum cycle
It can be observed in Figure 9 that compared to initial measurement (t=0h) there was an
appreciable change in DFR response in first six hours (t=6h) of vacuum cycle after the hot oil circulation.
The difference between the two curves is attributed to different temperature and high rate of moisture
extraction in the initial phase of vacuum cycle. After 6 hours, for the next 18 hours there was a uniform
change in DFR response every 6 hours indicating steady reduction in moisture concentration.
Page 56 of 284
The results in the next 30 hours of vacuum cycle as shown in Figure 10 did not show any
appreciable difference in the dielectric response of the system indicating a reduced efficiency of the dry
out system after first 24 hours. Utilizing the DFR data it could easily be concluded that the vacuum cycle
can be stopped as gain in moisture extraction compared to prolonged exposure of vacuum to winding
insulation would be of limited benefit. Once the system reaches low efficiency point, it is recommended
to perform another hot oil circulation and initiate a second vacuum cycle.
In this case, a second vacuum cycle did not show any major changes in dielectric response either
and it was decided to stop the process. Final moisture at the end of whole process was estimated to be
2.6%. Taking into account the age of the transformer and its operational history, final moisture levels
were considered to be in acceptable range by engineers and field operators. Service aged transformers
degradation of insulation over its life cycle cannot be reversed and moisture cannot be brought back to
very low levels as desired. Optimizing the dry out process is critical as many aged insulations cannot
sustain a long vacuum hold time that could impact the tensile strength of the insulation and can cause
serious damage to the insulation.
DFR technique could be used to determine the efficiency of the system and help in deciding the
opportune time to finish the process. It can also indicate the condition when no further moisture can be
extracted from the solid insulation without compromising the integrity of the insulation system.
Figure 11: Dry out data of a 765 kV 100 MVAr single phase reactor
Page 57 of 284
As observed in Figure 11, vacuum cycle was carried out for close to 32 hours to take into
account more than 100 microns change in a six-hour time period. At the end of vacuum cycle and just
before the final oil fill vapor pressure was recorded at 130 microns and winding temperature was
recorded at 60 ⁰C. Based upon piper chart information, moisture was estimated to be close to 0.15%.
Utility operators are always interested in percentage moisture in solid insulation after the final
oil fill to determine the transformer insulation condition before it will be commissioned. This embarks
the start of the life of the transformer. DFR measurement was conducted at the end of final fill on an oil
and paper insulation system and after a set time of 24 hours to allow for oil impregnation of the paper.
Figure 12: DFR response of a single phase reactor after the final oil fill
The insulation temperature was 26 ⁰C. The results are shown in the Figure 12 above. Percentage
moisture concentration was estimated to be 0.2 % of the weight of solid insulation. If possible, it is
recommended to perform the measurement at the end of one-week time period to ensure moisture
content is the same. The moisture estimated from vapor pressure and winding resistance data matched
very closely with the DFR moisture estimation.
Temperature dependence of the insulation system can be explored through DFR modeling. By
accurately taking the temperature effect into account, DFR modeling can be used to model the
temperature dependence of power factor over a wide temperature range based on measurements over
a frequency range [15]. From DFR technique the thermal profile of a dielectric insulation system can be
obtained. The thermal profile is shown by insulation power factor response as a function of
temperature. As discussed earlier, temperature, moisture and oxygen impact the life of a transformer
insulation the most, knowing insulation behavior at higher operating temperatures is of great value to
Page 58 of 284
the operators. Figure 13 below shows the comparison of thermal response before the dry out and after
the final oil fill.
Figure 13: Thermal response of dielectric insulation before dry out and after final oil fill
It can be observed that before the dry out (without oil), insulation power factor had a strong
temperature dependence especially at higher temperatures. However, dry out process was effectively
able to extract moisture from the solid insulation to below 0.25% limits and the insulation at the
completion of final oil fill had a very flat response to the change in temperature. It is to be noted that PF
values at 20 ⁰C did not show any significant change, this correlates well with the previous discussion of
PF not being a direct indicator of moisture present in the solid insulation.
Based upon the cases discussed in this section, it can be seen that DFR can be used to optimize
the process by knowing the initial moisture content before the dry out. The DFR data is then obtained at
every 6 hours interval during the dry out to identify the decrease in moisture content with time and find
out the efficiency of the process. Based upon the rate of decrease in moisture, decision can be made as
when to stop the process. At the completion of dry out and final oil fill, after a wait time of 24 hours
final measurement can be made to determine the reference moisture content in the solid insulation for
future trending during the life of a transformer or reactor.
Page 59 of 284
The correct time to stop the process based upon the rate of extraction and final moisture content is
a matter of debate. Optimizing this will prevent the exposure of insulation to prolonged hours of
vacuum and avoid any possible damage to the insulation.
Moisture estimation before and after the dry out are critically important for measuring the
effectiveness of the dry out system.
Dew point measurement only provides the average surface moisture content of the solid insulation.
A wait time of 12-24 hours should be allowed for equilibrium before the measurement is taken.
Power factor values do not directly correlate to the moisture present in the solid insulation.
Estimating moisture during the dry out using vapor pressure and winding temperature although
reliable will have its own sources of error.
Winding to winding insulation is recommended for DFR moisture estimation. In absence of that,
winding to ground insulation can be used for DFR measurement.
Running the DFR test during the vacuum process requires some setup that needs to be followed to
obtain reliable measurements.
Under vacuum it is not recommended to apply more than 50 V test signal for DFR measurement. In
an energized EHV station, lower frequency measurements may get influenced by static noise
present in the system.
Because of high winding temperatures during the dry out, DFR test can be conducted down to a stop
frequency of 5 mHz only for correct moisture estimation. This allows test to be completed faster, in
about 8 mins.
It is recommended to measure DFR every 6 hours during the vacuum cycle. However, measurements
made at night can be influenced by the condensation and humidity present in the atmosphere.
Moisture estimation using DFR during the dry out requires preliminary information about the
transformer or reactor for correct modeling, knowledge about the setup and factors that can affect
the measurements.
DFR during the dry out can be utilized to understand the efficiency of the process and helps in
making a decision as to when to stop the process based upon the gain with respect to vacuum time.
When performing the DFR measurement after the final oil fill on an EHV transformer or reactor, it is
recommended to use high test voltage signal (1400 V) to obtain more accurate measurements.
Page 60 of 284
The DFR application proposed in the paper can be used in improving the efficiency of the field
dry out process. It can help in determining the stage in the dry out operation where moisture extraction
process is no longer effective. This allow operators to decide if another heat and vacuum cycle should be
carried out or stop the process as further moisture extraction would be difficult to achieve without
affecting the integrity of the insulting system. DFR also provides thermal profile of the insulation system
indicating the temperature dependence and insulation behavior at those operating temperatures. As
described in the paper, DFR should be performed carefully, requires preliminary information of the
system to be tested and can be subjected to some of the field limitations during the vacuum cycle. From
cases described, it can be seen that DFR could be used as a powerful tool before, during and after the
dry out process to increase the confidence of operators by accurately knowing the percentage moisture
concentration in the solid insulation and hence, the condition of the insulation before transformers are
put in service.
VII. References
[1] “Field drying of transformer insulation Siemens”, Smart Grid Solutions
[2] “Dissolved Gas Analysis in transformer maintenance”, Engr. Inigo V. Escopete, Jr., AP renewables Inc.
[3] “Transformer maintenance: Facilities instructions, standards and techniques.” US Department of the
interior Bureau of Reclamation, Denver, CO
[4] “Oxidation Inhibitor and Reinhibiting Oil-Filled Transformers”, Andy Skholnik, Technical Brief, SD
Meyers.
[5] IEEE Std. C57.93-2007 “IEEE Guide for Installation and Maintenance of Liquid-Immersed Power
Transformers”.
[6] IEEE Std. C57.152-2013 “IEEE Guide for Diagnostic Field Testing of Fluid-Filled Power Transformers,
Regulators, and Reactors”.
[7] IEEE C57.106-2015 “IEEE Guide for Acceptance and Maintenance of Insulating Mineral Oil in Electrical
Equipment”.
[8] “Preventing Transformer Winding and Insulation Failure: SFRA and DFR”, Dinesh Chhajer, Weidmann
Annual Technical Conference, 2012.
[9] “IDAX User’s Manual”, 2008, PAX Diagnostics AB, Megger, Taby Sweden.
[10] “Final assembly, oil fill and processing specification for large power transformers.” American electric
Power.
[11] “Field processing methods for transformer dry out”, Weidmann Annual Technical Conference 2012,
Derek Baranowski, Baron USA.
Page 61 of 284
[12] “Theoretical Vapor Pressure Equilibria for Cellulose, air and oil”, Piper chart Source: American
Electric Power.
[13] “Transformer dew point measurement and insulation moisture estimation”, Part 1 & 2. Siemens
Smart Grid Solutions.
[14] “Continuous Monitoring of Power Transformer Solid Insulation Dry-out Process – Application of
Dielectric Frequency Response” Electrical Insulation Conference 2013, Diego Robalino.
[15] “Dielectric frequency response measurements and dissipation factor temperature dependence”
Electrical Insulation (ISEI) Conference, June 2012, Peter Werelius, Matz Ohlen, Diego Robalino, Jialu
Cheng.
VIII. Acknowledgment
The authors gratefully acknowledge the contributions of Dave Bush, Station Supervisor,
Transmission Field Services (American Electric Power) and Dr. Diego Robalino, Principal Engineer,
Technical Support Group (Megger) to this paper.
IX. Authors
Dinesh Chhajer received his Master of Science in Electrical Engineering
from University of Texas at Arlington. He is the Manager of Technical
Support Group at Megger USA. His responsibilities include providing
engineering consultation and recommendations in relation to testing of
transformers, circuit breakers and other substation assets. He has
presented a number of white papers related to asset maintenance and
testing at various conferences within power industry. He has previously
worked as an applications engineer at Megger and substation and
design engineer at Power Engineers Inc. He is currently a licensed
professional engineer in the state of Texas.
Page 62 of 284
Utilizing Visual and Thermal Video Analytics to
Enhance Operational Efficiency
Systems With Intelligence Inc.
E. Sotter, Ph.D.
P. Fernandez
R. Harada
Page 63 of 284
Utilizing Visual and Thermal Video Analytics to Enhance Operational Efficiency
E. Sotter, Ph.D., P. Fernandez, R. Harada
Systems With Intelligence Inc.
Abstract: Gone are the days of dedicated live view video screens that are manually monitored by
personnel. Advances in communications and software technology allow utility operators to be alerted of
events captured by these cameras as they happen. While physical security applications have seen the most
adoption of these video technologies within electric substations, visual and thermal monitoring can be
automated to run in a continuous manner to improve the reliability of the substation operation.
Real-time alerting and alarming of intruders within a substation, an impending equipment failure,
operation, or arc flash event directly to an operator or dispatcher, can take the most effective action.
Continuous thermal monitoring automates the manual “hand-held thermal scan” process that many
utilities practice. Utilizing thermal cameras systems with sophisticated video analytics algorithms for
anomaly detection can trigger automated alarms of possible issues directly to an operator who can make
an informed operational decision.
I. Introduction
Automated video systems provide utilities with real-time visualization of the substation allowing them to
make key decisions designed to prevent outages, save time and reduce costs. Substation video systems
use sophisticated multifunctional sensors in conjunction with video and thermal analytics to provide real-
time information for input into substation maintenance and operations programs. Using the latest
technology in video automation, these systems monitor key points in the substation providing visual and
thermal imaging that ties into the utility visualization and operation system (e.g. SCADA). Operators are
provided real-time visualization of the substation assets and are automatically notified when normal
operating conditions are not being met.
Substation monitoring systems differ from conventional video systems in function and application. A
substation monitoring system is designed to give utilities a real-time view of the operational status of high
value assets at remote substations from both the visual and thermal perspective. This differs from the
conventional video system that is focused on physical security. A substation monitoring system can
provide both operational and security functions [1].
The data obtained from video and thermal sensors deployed throughout the substation is processed by
an analytics engine in a server located on premise at the substation. These systems monitor the condition
on site and determine if an alert should be sent to the operations and control center. A typical substation
can have several video and thermal sensors deployed and it is not possible for operators to constantly
monitor them all, especially when multiple substations are involved. Traditional video systems require
Page 64 of 284
continuous manual human monitoring of a live stream of video. Video and thermal analytics automate
the monitoring process and provide the alerts directly to utility operational software systems, such as
SCADA, when further action is required. Operators can open the alert message to view a video snap-shot
of the event and open a real-time video feed from the substation to view the live scene (Figure 1). All the
video and alerts are recorded at the substation for review and archiving.
Visual monitoring has multiple operational, maintenance and safety applications in a substation that can
be done without sending a crew to the site:
Operational/Maintenance:
- Automated detection of events such as arc flash and switch arm movement
- Confirmation of remote controlled operations such as switch opening/closing
- Detection of animals around lines, switches, transformers
Safety:
- Ensuring safe conditions for the public and work crews at remote sites
- Witness and record that proper operational procedures are followed
- Intrusion detection, tampering, vandalism
To use the input from all video cameras installed in all substations for the tasks mentioned above, a
constant monitoring is required. However, simultaneously monitoring the video channels from all
substation is an impossible task for one human operator. Here is where the use of video analytics becomes
crucial.
Video analytics have been used for security video systems for many years. Most of these analytics utilize
algorithms that subtract the picture background from a video feed then analyze the shape and behavior
Page 65 of 284
of the objects left in the foreground. All these techniques have been mostly used to create video analytics
routines for security purposes (e.g. motion detection, virtual tripwires, loitering alarms, etc.) [2], however,
using the same concepts and algorithms, analytics can be manipulated to perform functions that are
specific to electric power applications. For example, if the analytic is fine tuned to a specific area and type
of movement, the analytics can determine the position of a disconnect switch arm and if it has moved.
Similarly, analytics can be tuned and focused on the switch contacts to detect an arc flash event that can
provide an operator information about an unintentional event or operation. Problems in a substation like
a fire or explosion can also be detected with this type of analytics (Figure 2).
Figure 2. Explosion due to a lightning arrester failure captured by a substation video monitoring system
For several years, infrared thermographic inspection has played a key role in maintenance plans in
electrical utilities due to its non-invasiveness, safety and relatively low-cost approach [3]. This technology
detects and measures infrared (IR) radiation coming from objects in the field of view of the thermal
camera and turns it into a visual image that represents the temperature of the objects. Traditionally, IR
thermographic inspections are done by technical experts using handheld thermal imaging cameras
targeting specific components in a substation like transformers, breakers, capacitor banks, control
cabinets, cable joints and insulation, etc. Due to the manual nature of the inspection, it cannot be done
continuously and it is very common to have several months pass between two consecutive thermal
inspections at a same site. Even though it is still a helpful tool to locate potential issues with components
in a substation, the prolonged period between manual thermal inspections makes it impractical for real
preventive and predictive maintenance. Furthermore, issues can arise very quickly in an electrical
substation causing considerable damage before a scheduled inspection takes place. Similarly, to a
permanently installed video security system, a fully automated thermal monitoring system that can run
continuously in a substation would be the adequate tool to support a preventive and predictive
maintenance plan.
Page 66 of 284
Every component or equipment in a substation has a normal temperature range of operation.
Components in the path of an electric current will dissipate power in the form of heat proportional to its
resistance to the current. Electromagnetic induction generates heat in the core of transformers which is
usually dissipated with radiator fins. Insulators on the other hand are not in the path of current, so they
should be at ambient temperature if they are in good condition. Understanding the temperature behavior
of critical components in the substation is crucial to establish the targets for an automated thermal
system: It is not always about finding the hot spot, but also the not-hot spot.
Here are some examples of situations with critical components in a substation that can be detected using
thermal cameras:
- Primary and secondary bushing connections: Loose or dirty connections, unbalance and
overloading problems can cause overheating in the connections (Figure 3). The connection with
the problem will be hotter than the connections of other phases.
Figure 3. Transformer showing center primary side bushing with external connection problem [4]
- Cooling fins: Cooling fins normally appear warm as they dissipate the heat generated in the core
and windings of a transformer. If it is seen that one or more fins have lower temperature than the
other fins (Figure 4.), then it can be concluded that the circulation of oil is being limited in those
fins, probably due to low oil level, flow obstruction, a closed valve, or perhaps the apparatus is
out of level [5,6]. An automated thermal system can detect an increment of the delta temperature
between fins and report it to an SCADA system.
Page 67 of 284
Figure 4. Transformer showing a problem with a cooling fin [6].
- Internal problems in bushings: A study found that, out of the problems a transformer can have,
bushing failures are the ones that most often lead to fire in a substation [7]. Thermal imaging can
show internal defects or connection problems in the bushing (Figure 5). It is important to
remember that the temperature inside the bushing will be much higher than surface temperature
detected by the camera.
Figure 5. Transformer showing left and right primary side bushings with internal problems [8].
b. Circuit breakers: Several different problems with circuit breakers can be detected by using
thermography (Figure 6). Like transformers, a faulty or dirty connection between bushing and
conductor cable or a bushing with structural damages will overheat during operation, which can
be detected with a thermal camera.
Page 68 of 284
Figure 6. Circuit breaker with a problem on the test tap cone [8]
There are two common ways to assess the thermal condition of components in a substation. One is
quantitative, where the objective is to get a fairly accurate temperature of the equipment. The other way
is qualitative, where the goal is to find difference in temperature between two components that should
have a similar temperature, which can be understood as a hot or cold spot [3]. Each technique has pros
and cons, and should be used depending on the type of automation process that is required. A good
thermal camera system should provide tools to use both methods.
a- Quantitative method
In this method, the temperature readings from the IR camera are used as the real temperature of the
component. This temperature reading can then be compared to a threshold set by the component
manufacturer for normal operation, and trigger alarms when the temperature passes that threshold. It
can also be used to create models that can help to predict when a component may break.
The problem with this method is the factors intrinsic to the component in which temperature is being
measured (i.e. emissivity, size and distance from the camera), as well as environmental (i.e. ambient
temperature, humidity, rain, wind, velocity and solar radiation), that affect the IR light reaching the
camera [9]. All these factors should be taken into account when the temperature reading is calculated to
have a valid useful temperature value. Factors related to the component can be provided by the
manufacturer or measured directly in the substation. Some electrical substations have weather stations
that can be used to provide the required environmental inputs to the thermal system. However most of
the times these factors are unknown. This is when the qualitative method is used.
b- Qualitative method
In this method the temperature readings from two components, whose difference in temperature or delta
temperature (∆T) is known, are compared. If the ∆T obtained from the measurements is different than
the expected ∆T then an alarm is triggered. For this method to work, the components that are being
compared should be made of the same material and be at a similar distance from the IR camera. As both
Page 69 of 284
components are affected by the same factors, those factors can be ignored from the calculations and still
obtain a fairly accurate measurement of the ∆T between the two components.
This method can easily detect problems in the substation, however it cannot be used for modeling and
failure prediction as it does not guarantee an accurate reading of the temperature of the component.
IP networks are commonly deployed in substations and other industrial applications to connect Intelligent
Electronic Devices (IEDs), Remote Terminal Units (RTUs), and other protection and control equipment to
each other and to the SCADA system [10,11]. Substation automation networks are industrial, process
based networks that run critical applications to keep substation equipment and the power system
protected and running safely [12]. Process based communications is machine to machine so uninterrupted
and timely delivery of data is critical to keep processes running correctly. Process based networks must
therefore be kept isolated from the Internet and other corporate traffic to ensure that the data flow is
secure and free of losses. In an IP network, technologies such as Virtual Private Networks, (VPNs), subnets
and firewalls are used to keep the substation network isolated from the rest of the corporate LAN [13].
The video monitoring system is connected on the substation network so it can communicate with the
SCADA system at the control center with alarms, messages, visual and thermal information of the
operating conditions at the substation. The information can be fed into the control system for automated
responses or for operators to make decisions and take actions. Since the substation network requires
access privileges, the video and SCADA systems are not accessible to personnel without proper
authorization.
Representation of an electric power system traditionally consists of a one-line diagram that shows
essential information such as current and voltage levels, switch status etc. in a two-dimensional map. The
information is real time and shown on the Human Machine Interface (HMI) of the utility SCADA system.
Integrating a video system with SCADA adds another dimension of real time visualization critical in
ensuring more informed operational decisions to minimize outage times when unavoidable, while also
providing trending analysis input to prevent outages and prolong the life of high value assets.
While a SCADA or OMS/DMS system can provide critical information on the status of the system such as
a breaker trip and power outage, visual and thermal video information can provide confirmation of the
cause of the event and enhance real time situational awareness. Visualization may be able to provide
insights as to why the breaker operated? What is happening now at the remote site? Is a power line down?
Is there a fire or leakage? With this visual information, the utility can safely and securely reduce the outage
time by having knowledge of the problem before rolling the truck and sending personnel into the area.
More importantly, video and thermal information can reveal system inefficiencies by finding temperature
hotspots in the system so the utility can make repairs before the failure occurs. Transformer losses,
voltage drops and load imbalances can be caused by leakage current in insulators or high resistance
connections in power system components. These components will often show increased heat buildup that
Page 70 of 284
can be detected by monitoring with thermal cameras. Continuous monitoring will provide alerts and
trending data as the system goes through load cycles allowing the utility to perform condition based
maintenance on the system before failures and outages occur.
VIII. Conclusion
Utilities are faced with the challenge of maintaining/improving service reliability without increasing costs.
Video automation can remotely predict failures before they happen and detect and diagnose problems
when they do. Video analytics automate the monitoring process to minimize the workload on operators
and provide visual confirmation of the conditions at the substation. Through (1) automated remote visual
inspections that reduce the number of trips to remote sites and (2) continuous thermal monitoring that
analyzes the operating temperature of components in the power system driving condition based
maintenance activities, automated visual and thermal monitoring can enable more efficient electric utility
operations
IX. References
[1] E. Sotter, G.P. Porciello and J. McClean, Video monitoring solutions for electric utilities: Issues,
requirements and examples, Electric Energy T&D Magazine, Jan-Feb, 1 (17) (2013) p.39-43
[2] S. Velastin, CCTV video analytics: Recent advances and limitations, Lecture notes in computer
science, Vol 5857 (2009) p.22-34
[3] M. Jadin and S. Taib, Recent progress in diagnosing the reliability of electrical equipment by using
infrared thermography, Infrared Physics & Technology 55 (2012) p236–245
[4] S. James, Common thermography uses and applications within the petrochemical, offshore oil and
gas, chemical, and power generation industries. http://www.irinfo.org/02-01-2009-james/
[5] A. Huda and S. Taib, Application of infrared thermography for predictive/preventive maintenance of
thermal defect in electrical equipment, Applied Thermal Engineering 61 (2013) p220-227
[7] F. Vahidi and S. Tenbohlen, Statistical failure analysis of european substation transformers, ETG-Fb.
144: Diagnostik elektrischer Betriebsmittel 2014, p.5-9
[8] M. Ralph, Power plant thermography–wide range of applications, Inframation 2004, Proceedings vol.
5, p241-258
[9] L. Dos Santos, Infrared thermography applied for outdoor power substations, Thermosense XXX Proc.
of SPIE Vol. 6939, 69390R, (2008)
[10] K.H Mak and B. Holland, Migrating electrical power network SCADA system to TCP/IP and Ethernet
netwroking, Power Engineering Journal, 16(6) (2002) p.305-311
Page 71 of 284
[11] C. Wester and M. Adamiak, Practical applications of Ethernet in substations and industrial facilities,
Conference Record of Annual Pulp and Paper Industry Technical Conference,(2011) p.55-66
[12] C. Hoga, New Ethernet technologies for substation automation, 2007 IEEE Lausanne Powertech,
Vols 1-5 (2007) p.707-712
[13] S. Wu, C. Kiu and A. Stefanov, Distributed specification-based firewalls for power grid substations,
IEEE PES Innovative Smart Grid Technologies Conference Europe (2014)
Page 72 of 284
DESIGNING GROUNDING GRIDS
BY
Chief Engineer
And
Page 73 of 284
Designing Grounding Grids
Abstract:
Modeling/designing existing or new grounding grid is a challenge for designer. Grounding grids
usually consist of ground conductors, ground rods, and/or UFER grounds. The various
grounding elements are installed at different depths. Soil resistivity may vary significantly
throughout the site. This presentation will discuss the modeling of two large fossil fuel
generating stations in Texas. One generating station was constructed with zinc ground rods
and insulated conductors for connecting the rods. The other was constructed with UFER
grounds, copper clad ground rods, and bare copper grounding conductors. This paper will
Introduction
Electrical grounding system serves many purposes. It adequately relays the fault. An effective
electrical ground must have a low enough Resistance-to-Remote Earth to assure an essentially
instantaneous trip to the largest breaker in the facility. Effective grounding and bonding of
cabinets, enclosures, and other structural components assures personnel safety when
(ESG) assures safe step, touch, and mesh potentials throughout the facility. Signal to reference
grounds are essential for reliable operation on modern computerized Distributed Control
Systems (DCS).
Page 74 of 284
Case History 1:
A large, three-units lignite-fired steam electric station was designed and constructed in East
Texas with a total capacity of 1880 mW. Unit 1 and 2 are each rated 565 mW and constructed
in 1974 and 1976. Unit 3 is rated at 750 mW and constructed in 1978. The purposes of this
study were to evaluate the present condition of Electrical Safety Grid and make
recommendations for repairs or upgrades that might be necessary for adequate relaying and to
Underground utilities at the plant site include, at a minimum, the following components:
Compressed air lines, service water lines, DI pipe, firewater, and potable water lines that may or
may not be coated. Bar grilles, traveling screens, and circulating water pumps at the intake
structure. Dual prestressed concrete pipe circulating water lines, various control lines, electrical
conduit, and foundation rebar. An extensive bare-copper grounding grid containing 500-MCM
Point-to-point (PTP) resistance measurements were made between critical grounding points
throughout the facility. During the course of this study 436 test points were established. PTP
resistance measurements of 0.002 ohm, or less, are taken as a clear indication of effective
electrical continuity, not necessarily requiring calculation of the theoretical PTP resistance. For
resistance measurements to a higher value, the theoretical PTP resistance for a single Number
4/0 AWG or larger conductor for the circuit path is calculated. If values fall within 110 percent of
the theoretical value, adequate electrical continuity is also indicated. Many of the measured
values fall far below the theoretical resistance for a single conductor of measured length due to
the presence of multiple parallel current paths. Some of these alternative current paths may
Page 75 of 284
include underground metallic piping, such as fire lines within the facility. Metallic piping in never
accounted for in grid design calculations. Due to the presence of mechanical fittings in piping
runs, such as valves or bell and spigot joints, adequate electrical continuity is not always
assured. One hundred forty-seven (147) measurements or 33.3 percent failed, indicating
somewhat excessive resistance between those particular points. Most of the failures were tests
to fire hydrants (mechanical fittings) and a few were to isolated lighting standards. Cut, missing,
or broken short conductor segments within the interior of a large grid may be safely tolerated
since mesh voltages are expected to be much lower than for similar sized meshes at the
perimeter of the grid. However, it is not safe to make such as assumption without careful
modeling of the grid. In each case, we determined that the lighting standards were adequately
The results of our etap modeling efforts showed a basically safe grid with the unsafe regions, for
touch potential only, confined to perimeter areas of the grid. An inspection of the etap plots for
step potential and for touch potential confirmed very low mesh potentials within the interior of
the grid. Therefore, the few cut, broken, or missing conductors identified within the grid interior
For electrically unsafe grids, or small segments of a grid, any one or a combination of the
following techniques may be employed to assure personnel safety. Add grid conductor,
increase the area of the grid, install additional groundrods, include interior perimeter fence
grounding and/or groundrods with the internal grid, install additional perimeter conductor fence
grounding outside of the fence, install or improve surface treatments, reduce mesh size/area in
limited areas of the grid where unsafe step or touch potentials are shown to exist.
Assuming an adequately low grid resistance to meet system standards, our first preference
Page 76 of 284
would be to employ surface treatments for personnel safety to small local areas where unsafe
touch potentials might exist. In each instance, we have recommended the lowest cost safe fix
that also has the least adverse effect on system operation during construction of the
modifications.
We measured 0 – 2.5 foot soil resistivity measurements at multiple locations within the plant
boundaries, these data are used for the step and touch voltage calculations. In the interest of
assuring adequate electrical safety throughout the ESG, we utilized a value of 18.6 Ohm/meter
in the calculations, which was the lowest measured value for surface soil resistivity within the
facility. This results the calculated safe step and touch voltages for a 50-kg body weight using
the lowest measured resistivity of 309.6 V for safe step potential in bare earth, 267.2 V for safe
touch potential in bare earth. These values make no allowance for surface material (crushed
rock or pavement) to increase human contact resistance to earth, but assume a bare earth
model throughout the grid (based upon 0-2.5 foot soil resistivity measurements) for calculation
of tolerable step and touch voltages. For the case of 3 inches of asphalt, 506.0 V is a safe
touch potential. Maximum clearing times are conservative estimates based upon best available
This site contains a very large ESG which initially stretched the finite element component of the
etap grid module beyond its limit. We made multiple runs, gradually changing default settings
within the program. It was necessary to change the default setting for the finite element length
from three (3) feet to ten (10) feet. Other simplifications were also necessary. One of the runs
required seven (7) days to reach a solution using an engineering work station with a very fast
CPU. For the initial grid in question, as found, the etap model calculated a resistance-to-remote
earth of 0.02 ohm for the entire ESG. For the given maximum system fault current, this results
in a grid potential rise (GPR) of 926.4 Volts. Maximum calculated step potential for any mesh
Page 77 of 284
within the grid is 121.0 Volts which is only 39.1% of the calculated safe limit. However, touch
potential was another matter. Maximum calculated touch potential for any mesh within the grid
was 502.3 Volts which was 88.0% higher than a safe value.
A careful assessment of unsafe touch potential regions within the grid, or its perimeter, identified
many areas where there are no grounded structures with which to come in contact. Therefore,
no remediation was required at that time. We identified five (5) areas of the grid where plant
personnel could come into contact with a grounded structure that would be unsafe under a
system fault condition. Two (2) of those areas did require the addition of additional 4/0 grid
conductor. One of those areas required two (2) additional ground rods. In the other three (3)
cases it was far more economical to provide some form of surface treatment than to install very
significant lengths of additional grid conductor. Asphalt paving met that requirement in each of
those areas. The limited grid additions did result in slight improvements in the form of reduced
GPR and mesh potentials. GPR was reduced from 926.4 Volts to 917.9 Volts and maximum
touch potential was reduced from 502.3 Volts to 492.8 Volts, a 1.89 percent reduction.
Case History 2:
A 1025 mW capacity large, gas-fired steam electric station was designed and constructed in
Houston, Texas in 1960. Some real estate of the station was sold to other business and
removed from original construction of the plant. This activity was partially driven by grid
discrepancies that were unearthed when troubleshooting damaged header cables in a cathodic
protection rectifier groundbed. Rectifier current output significantly declined in 2010 when the
current output fell to about 15 percent if its previous value. Subsequent exploratory repairs
revealed serious damage to the header cables in the groundbed at multiple locations. It is
speculated that grounding grid conductors were cut at the same time, rendering the cooling
Page 78 of 284
tower grid isolated from the main plant grid. Therefore, it was necessary to turn off the rectifier
in order to avoid serious corrosion to sections of the grid and buried plant piping (BPP) due to
cathodic interference. This unit was returned to service after the appropriate grid repairs were
made. The purposes of this study were to evaluate the present condition of the ESG and to
make recommendations for such repairs or upgrades that might be necessary to assure
adequate relaying and to assure personnel safety plant wide. Here the grounding grid was
constructed with insulated copper conductors and bare zinc anodes serving as ground rods.
During the course of this study 237 test points were established. Point to point (PTP) resistance
measurements criterion were the same as for case 1. There is a small amount of larger 500
MCM conductor installed around transformers or other high current areas of the plant. In those
instances, resistance calculations were performed with the appropriate linear resistance value.
Whenever the second etap© grid model showed unsafe touch potentials, the appropriate grid
The testing revealed serious electrical discontinuities in the ESG requiring immediate repair.
Seventy six (76) measurements or 32.76 percent failed to show adequate electrical conductivity,
indicating excessive resistance between those particular points of measurement. At five (5) of
the locations tested we found broken grounding pigtails aboveground that also required repair.
We measured 0 – 2.5 foot soil resistivity at multiple locations within the plant boundaries, these
data are used for the following step and touch potential calculations. In the interest of assuring
adequate electrical safety throughout the ESG, we utilized a value of 10 ohm/meter in the
calculations, which was the lowest measured value for surface soil resistivity within the facility.
Under a system fault, for a 50-kg body weight using the lowest measured resistivity for the bare
Page 79 of 284
earth condition, the safe step potential was calculated as 445.9 V and safe touch potential was
468.1 V.
Higher tolerable Voltages would apply for an individual standing on asphalt or concrete
pavements, such as within the power bock. But we do not design to those conditions when the
majority of the plant is defined by the bare earth condition. Therefore, these values make no
allowance for surface material (crushed rock or pavement) to increase human contact
resistance to earth, but assume a bare earth model throughout the grid for calculation of
tolerable step and touch voltages. Maximum clearing times are conservative estimates based
upon best available information provided to us by client. We made multiple etap© runs, to
simulate the various conditions encountered. The following individual conditions or case studies
were modeled:
Grid 1 – The entire plant grid as originally designed, board drawn, and constructed. The model
shows a safe grid with the highest calculated touch potential of 365.4 Volts or 80 percent of the
safe limit. Maximum calculated step potential was 137.08 Volts or only 29.5 percent of the safe
limit. For the initial grid in question, as found, the etap© model calculated a resistance-to-
remote earth of 0.01 ohm for the entire ESG. Under the given maximum system fault current,
Grid 2 – The plant grid was originally designed and constructed with a contigous grid within the
fuel oil tank farm. During a change of ownership, the tank farm was sold to others and the tank
farm grid was no longer connected to the plant grid. When designing an ESG, the initial
Page 80 of 284
parameter, and generally the most significant, is the surface area covered by the grid. By
removing the tank farm bottoms, fuel piping and ESG elements within the tank farm, the surface
area of the grid was reduced by about 37.5 percent. This resulted in a nineteen (19) percent
increase in grid resistance-to-remote earth and, correspondingly, GPR under fault conditions.
The model clearly showed an unsafe grid with the highest calculated touch potential of 462.3
Volts or 103.7 percent of the safe limit. This touch Voltage is 26.5 percent higher than was the
case for the original grid. Maximum calculated step Voltage is only 34.7 percent of the safe
limit. This is an unintentional consequence of selling the tank farm to another entity and
disconnecting fuel piping and grid conductors from the main plant grid. Therefore, some of the
eastern side of the plant yard was unsafe for human contact under fault conditions.
Grid 3 – The plant grid as configured with all identified electrical discontinuities modeled. This
model also showed an unsafe grid with the highest calculated touch potential at 103.6 percent
of the safe limit. Maximum calculated step potential is only 34.7 percent of the safe limit.
Grid 4 – The cooling tower (CT) grid that was isolated from the main plant grid the cooling tower
area was very unsafe for human occupancy. The estimated maximum fault current and clearing
time was derived from the I2t curves published by GE for the Type AK2A-25-1 Low Voltage
Power Circuit Breakers used to protect the circulating water pumps at the cooling tower. When
the three grid stingers from the CT to the power block were intact the vast majority of the fault
current would return to generation via the three stingers resulting in very little local earth GPR.
With all the three cut stingers, all of the fault current must return to generation via discharge
from the CT local ESG grounding electrodes. The etap© model for a fault at the CT with three
cut stingers showed an extremely unsafe CT grid with the highest calculated touch Voltage at
Page 81 of 284
635 percent of the safe limit. Maximum calculated step Voltage was also unsafe at 350 percent
of the safe limit. Therefore, we could expect to have a fatality for anyone in direct contact with
the structure or possibly even if standing close to the CT should one of the circulating water
pumps fault to ground. Plant personnel were directed to barricade off the entire area. Entry
was not permitted with the pumps energized until grid repairs were completed.
Grid 5 – The entire plant grid as constructed after the required repairs were completed. The
etap© model shows a safe grid with the highest calculated touch potential of 331.8 Volts or 74.4
percent of the safe limit. The maximum calculated step potential is 171.4 Volts or only 37.2
percent of the safe limit. There are many other repairs that could have been made, but the
recommendations contained in the report provided the minimum effort that was necessary in
order to assure a safe grid, plant wide. An inspection of the etap© plots for step potential and
for touch potential confirmed very low mesh potentials within the interior of the grid. Therefore,
most of the few cut, broken, or missing conductors identified within the grid interior by this study
did not require repair or replacement. Repairs to existing grid conductors were made with like
kind ie. Number 4/0 AWG insulated copper conductor. However, for the additional grounding
required in order to assure a safe grid, we made a better conductor choice. We recommend the
installation of bare galvanized steel ½ inch diameter wire rope. This material is much less costly
than copper, is fully compatible with the plant wide Cathodic Protection system, and has the
added benefit of providing additional grounding to supplement the few proposed additional
ground rods that were required. This significantly reduced the number of additional ground rods
Page 82 of 284
Conclusions:
Modeling of two large grounding grids has been discussed. One grounding grid solution was
provided that did not make the whole plant area “safe” for touch potential. The objective was
making the plant area safe for Touch Potential where human contact was possible. The
solution was provided by adding additional conductors, ground rods, or surface treatment where
necessary. The other grounding grid solution was provided by adding galvanized aircraft cable
grounding conductors instead of insulated ground conductors and supplemented with bare
anodes in special backfill. This conductor can use where plant is cathodically protected. This
These two case histories clearly document that electrical grounding cannot be taken for granted,
assuming that all is well. If grid resistance-to-remote earth is too high, relays will not trip out as
clearing times can render an otherwise safe grid unsafe for human occupancy. And an
Page 83 of 284
Transformer tank vibration model to
detect winding deformations
By
Page 84 of 284
Abstract
Twelve to fifteen percent of transformer failures are caused by winding deformation due to the
high electrodynamic forces that appear during through faults. These geometric variations lead to
a pattern change and an increase of the winding vibration and, consequently, to an increase of the
solid insulation mechanical fatigue, the insulation then can be degraded and short circuits
between turns will show up. These winding deformations can also change the distance between
conductors changing its dielectric withstand capability as well as the windings series and shunt
capacitances and, thus, the voltage distribution in case of lightning or switching over voltages,
making it different from what the transformer was designed to withstand for, increasing the risk
of failure. These considerations make clear the relevance of an early detection of winding
deformations.
The vibration model will determine the vibration pattern of the transformer tank as a way to
detect transformer winding deformations, and it is proposed as a complementary technique to
Frequency Response Analysis (FRA) or Leakage Reactance Measurement (LRM) having the
advantage that it can be used for online monitoring and, thus, catastrophic failures can be
avoided between successive maintenance outages.
Introduction
Vibration analysis is a key test in rotating machines’ predictive maintenance programs and is
widely used to detect on load tap changer failures by means of the noise signature analysis
during tap regulation. Use of the main tank vibration as a diagnosis tool is not very common,
nonetheless some references have been found as to the monitoring of this variable.
In this paper, the use of a model to monitor tank vibration developed in the University Carlos III
of Madrid [1] is presented. The strategy of model-based transformer monitoring systems was
proposed by MIT researchers [2] with the aim of allowing the early detection of failures in a
transformer. For this purpose, some transformer key variables are calculated by models from
some input variables. A great difference between calculated and measured values is an indication
as to a structural change in the transformer, and an alarm is then emitted. The advantage of
model-based monitoring in contrast to a traditional threshold-based one is that transformer global
working conditions are taken into account when emitting an alarm.
Vibrations in a transformer are generated by the different forces appearing in the core and in the
windings during the operation. The main sources of vibration are mainly voltage and load current
depend.
Voltage-dependent vibration
Page 85 of 284
change their shape when subjected to a magnetic field. The Weis Domains in metal align
themselves along the time-varying magnetic main flux induced by the applied voltage. The
changing magnetic flux density forces the Weiss Domains to follow the flux by rotation which
claims certain area in the material, their movement result in a changing length of the whole
material. Expanding and tightening lamination sheets causes mechanical vibration. The
mechanical orientation of the Weiss Domains at the positive and negative maximum of the
magnetic flux density is the same. One electrical period leads to two maxima of material
expansion. Therefore the basic oscillation is double the fundamental electrical frequency. Fig. 2
shows the relationship between length variation (in %) and iron magnetic flux density. As it can
be seen, the curve represents hysteresis. Neglecting the hysteresis effect, this curve (plotted in
the continuous line in Fig. 1) can be replaced by the idealized curve plotted in the dashed line in
Fig. 2. The mathematical expression of the idealized curve can be approximated to a quadratic
law, establishing a linear relation between the elongation and the flux density squared. Taking
into account the relation between applied voltage and flux density and admitting the elongation is
proportional to the force, the result is magnetostriction forces being proportional to voltage
squared
Load-dependent vibration
Load-dependent vibration is originated by the load current in the windings due to electrodynamic
forces caused by the interaction of the current in a winding with leakage flux. These forces are
Page 86 of 284
proportional to the current squared and they have components in axial and radial directions.
Axial forces tend to compress the winding vertically. In a simple case of a two-winding
transformer, radial forces tend to compress the internal winding and to expand the external one,
since currents in both windings flow in opposite directions. Fig. 2 shows the directions of radial
and axial forces and their relative magnitude depending on the height and the radius of the
transformer winding.
Vibration Model
Page 87 of 284
service transformers, all of the input and output variables are measured using external sensors
only.
The forces analyzed in the previous section give rise to vibrations in the core and in the winding
in various directions. Core and winding vibrations interact and transmit through the oil and the
transformer supporting elements to the tank. To establish the model equation, it is assumed that
the tank vibration in a given direction results from the addition of the winding and core
vibrations in that direction multiplied by corresponding transmission coefficients (1).
Taking into account that winding and core vibration are proportional to current and voltage
squared, respectively, and that both variables present their main components at 60-Hz frequency,
the result is (2).
To compute tank vibration from the superposition of core and winding vibration, the relative
phase angle between both vibrations must be taken into account. As winding and core vibration
depend on current and voltage, respectively, the phase angle between them depends on the power
factor of the electrical load being transmitted through the transformer. To consider the phase
angle between vibrations, the model is formulated in complex variables.
Temperature Influence
The vibration model algorithm is given by (2), where Vtank, i and u are complex variables that
correspond to the real and imaginary parts of the tank vibration, current, and voltage main
frequency components, , , , and are complex parameters whose value depends on
transformer geometry and which must be computed from data measured on the transformer and
Page 88 of 284
t is the oil temperature. The parameters in (6) change slowly in time as the transformer ages,
clamping pressure decreases, so vibration becomes greater. Because of that, parameters must be
recalculated after some years using the model.
The model described in this paper is based in the analysis of the periodic vibrations and does not
consider the LTC vibrations as they appear only after an LTC operation, producing transient
vibrations that are added to the periodic ones caused by the core and the windings. The elements
of the cooling system (i.e., oil pumps and fans) also generate vibrations that are added to the
main vibrations. These elements do not operate in a continuous way, as they are switched on or
off depending on the transformer working temperature, but when they work, a periodic vibration
signal appears, the effect of this vibration is considered by the model by a different set of model
parameters.
WEG Mexico allowed us to test the model in one of their transformers. A set of measurements
was performed on a transformer to adequately choose the point of the tank where vibrations
should be measured, so that it reflects the transformer internal behavior as well as possible. The
test transformer was fitted with external accelerometers and WEG’s test facility we were only
able to simulate no-load and load conditions one at a time. Disappointedly it was found out that
the model didn’t work out under no-load conditions because there was no load current, but it did
work under load conditions even with only impedance voltage present. Fig. 3 shows the vibration
model output compared to actual data taken during one of the test performed
Page 89 of 284
As mentioned it was not possible to simulate real operating conditions, therefore the next step is
to test the model in an actual transformer in service where voltage and load current will be
present at the same time so the model can be fully tested.
Conclusion
In this paper, tank vibration monitoring has been proposed as an on line method to detect
winding deformations in power transformers. The model was developed to calculate tank
vibrations, taking into account transformer operating conditions. The vibration calculated value
is compared with the measured one being a great discordance between both values a sign of a
possible change in the transformer geometry. The model takes into account how the vibrations
are generated at different transformer components. The model includes the effects of temperature
on vibration generation, superposition, and transmission to the tank. Only the main component of
tank vibration is used in the model, as higher harmonics are much lower and do not show a
strong dependence with current. Tests were carried out to localize the best point of the tank to
place the accelerometers to measure vibrations. The test results are promising but need to be
confirm a transformer under actual operating conditions.
References
Biography
Page 90 of 284
Steady-State and Dynamic
Transmission System Model
Validation (NERC MOD-033-1)
Prepared for
Prepared By
Page 91 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
TABLE OF CONTENTS
Page No.
1.0 ABSTRACT .......................................................................................................... 2
2.0 INTRODUCTION .................................................................................................. 2
3.0 POWER FLOW MODEL VALIDATION PROCESS ............................................. 2
3.1 Model and Data Preparation ....................................................................................3
3.2 Adjustment of Planning Case to Match Selected System Conditions .....................4
3.3 Solving Power Flow and Comparison of Key Quantities ........................................4
4.0 DYNAMIC MODEL VALIDATION PROCESS...................................................... 4
4.1 Model and Data Preparation ....................................................................................5
4.2 Adjustment of Power Flow and Dynamics Cases ....................................................6
4.3 Simulation of Dynamic Event and Comparison ......................................................6
5.0 POWER FLOW MODEL VALIDATION RESULTS .............................................. 6
5.1 Model and Data Preparation ....................................................................................6
5.2 Adjustment of Planning Case to Match Selected System Conditions .....................7
5.3 Solving Power Flow and Comparison of Key Quantities ........................................7
5.4 Model Enhancement ................................................................................................8
6.0 DYNAMIC MODEL VALIDATION RESULTS ...................................................... 8
6.1 Selection of Event for Validation ............................................................................8
6.2 Model and Data Preparation ....................................................................................9
6.3 Simulation of Dynamic Event and Comparison ......................................................9
6.3.1 Bus Voltage Comparison (DFR Vs PMU Vs PSS/E) .............................10
6.3.2 Real Power (MW) Flow from McWilliams 115 kV Bus ........................10
6.3.3 Reactive Power (MVAr) Flow from McWilliams 115 kV Bus ..............10
7.0 MODEL VALIDATION RESULT ANALYSIS AND CONCLUSIONS ................. 11
7.1.1 Power Flow Model Validation ................................................................11
7.1.2 Dynamic Model Validation.....................................................................12
ACKNOWLEDGEMENTS ............................................................................................. 13
REFERENCES .............................................................................................................. 13
LIST OF TABLES
Page No.
Table 5-1: Total In-Service Load Comparison ................................................................................7
Table 5-2: Total Tie Flow Comparison ...........................................................................................7
Table 5-3: Branch Flow Comparison ...............................................................................................7
Table 5-5: Branch Flow Comparison ...............................................................................................8
LIST OF FIGURES
Page No.
Figure 3-1: Power Flow Model Validation Workflow ....................................................................3
Figure 4-1: Dynamic Model Validation Work Flow .......................................................................5
Figure 6-1: Location of Disturbance on April 05, 2017 ..................................................................8
Figure 6-2: Voltage Comparison (DFR Vs PSS/E) .......................................................................10
Figure 6-3: MW Flow Comparison (PMU Vs PSS/E) ..................................................................10
Figure 6-4: MVAr Flow Comparison (PMU Vs PSS/E) ...............................................................11
Figure 7-1: Alternative Tie-Line Flow Equivalent ........................................................................12
Page 92 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
1.0 ABSTRACT
The design, planning and daily operation of power systems depend on the accurate modeling of system
elements such as power plants, transmission lines and loads. The NERC (North American Electric
Reliability Cooperation) MOD-033-1 standard establishes new requirements for the validation and testing
of steady-state and dynamic planning models. PowerSouth Energy Cooperative has completed its first
round of NERC MOD-033-1 compliance studies: creating validation guidelines and metrics, and validating
the steady-state and dynamic planning models utilizing PSS ODMS State Estimator cases, and PMU
(phasor measurement unit) values. The compliance study solved many practical challenges within the
model validation process and revealed opportunities for improved modeling & testing in the future.
2.0 INTRODUCTION
The validity and usefulness of planning studies depends on accurate modeling of steady-state and dynamic
power systems. Inaccurate modeling can produce major discrepancies between simulated and actual system
responses, which will further lead to incorrect prediction of system behavior [1][2]. The North American
Electric Reliability Cooperation (NERC) developed the MOD-033-1 standard [3] to “establish consistent
validation requirements to facilitate the collection of accurate data and building of planning models to
analyze the reliability of the interconnected transmission system” and requires all planning coordinators,
reliability coordinators and transmission operators to comply, starting in 2017.
To comply with the NERC MOD-033-1 standard, Burns & McDonnell and PowerSouth Energy
Cooperative (PowerSouth) developed a set of procedures and guidelines for validation of PowerSouth’s
steady-state and dynamic planning models, and validated the current planning models against real-time
system conditions and actual system responses to a disturbance event.
Page 93 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
Figure 3-1: Power Flow Model Validation Workflow
Page 94 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
1. Generation MW Dispatch
2. Transmission topology
3. Station loads (MW, MVAr)
4. Switch shunt status/output
5. Transformer tap positions
Additionally, two model component states must be adjusted indirectly to bring better alignment between
the planning model and the state estimator case: tie-line MW flows and MVAr flows, and generator MVAr
outputs. To reconcile tie-line flows, artificial loads are placed at the boundary buses and adjusted until the
tie-line flows match the state estimator case. To reconcile the generator MVAr outputs, the unit terminal
voltage is adjusted until the outputs match the state estimator case.
The comparison results are recorded and utilized for planning model enhancements.
Page 95 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
Figure 4-1: Dynamic Model Validation Work Flow
Page 96 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
The system responses within the adjusted cases are plotted and compared to the real-time event recordings
(PMU and DFR). Unlike the steady-state model validation, the dynamic model validation is qualitative.
Meaning, the plots of actual system responses versus simulated system responses must be visually
compared rather than measured against discrete criteria. The following lists the aspects of the event that
should be compared.
The comparison results with major discrepancies will be analyzed and utilized for planning model
enhancement.
Page 97 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
The results show that the adjusted power flow planning model had only minor discrepancies with the real-
time snapshot on system level, thus is valid for further comparison.
Additionally, all MW and MVAr flows on the PowerSouth tie-lines and transformers were within ±10
MW/MVAr. All generator MW outputs matched the realtime values, MVAr outputs were with within ±10
MVAr and the terminal voltages were within ±3%.
Page 98 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
Page 99 of 284
Steady-State and Dynamic Transmission System Model Validation
(NERC MOD-033-1)
TECHNICAL PAPER
The transient disturbance event was recorded in Digital Fault Recordings (DFR) and Phasor Measurement
Units (PMU). The DFRs are located on various circuit breakers and PMU units are available in specific
115 kV buses within the PowerSouth system.
The 2018 off-peak power flow planning model along with a corresponding dynamic model was selected as
a baseline due to its dispatch and topology similarities to the April 5th, 2017 system conditions. Using the
process outlined in Section 3.0, the 2018 off-peak planning power flow model was adjusted to match the
system conditions at a specific time of the dynamic event.
The MVAr flow comparisons for the McWilliams to Victoria Junction 115 kV transmission line, and the
McWilliams to Wiggins Junction 115 kV transmission line, are shown in Figure 6-4.
The 2017 winter peak power flow planning model closely matched the characteristics of the system peak
conditions observed on January 8th, 2017. The recommended comparison criteria are as follows:
• Unit MW and MVAr output, generator voltage, and BES bus voltages should be within 3% of the
state estimator value.
• MW and MVAr flows on BES transmission lines, tie-lines and transformers should be within
10MW/MVAr of the state estimator value.
The following recommendations would enhance the future development of the power flow planning models
and adjusted power flow planning models to comply with NERC MOD-033-1 requirements:
1. Closely compare the transmission line, tie-line and transformer impedances between the
PSS/ODMS case and the planning model.
2. To mitigate tie-line flow discrepancies, consider replacing the tie-lines (within the adjusted
planning model) with two artificial loads (added at both ends of the tie-line) to force the exact tie-
line flow, as shown in Figure 7-1.
Artificial Load
Artificial Load
PowerSouth Area
Boundary Bus
The following recommendations would enhance the future development of the dynamic planning models
to comply with NERC MOD-033-1 requirements:
1. Audit the PMU measurements and time stamps to verify their accuracy, as the DFR and PMU
measurements during the study revealed a consistent time delay in the PMU measurements.
2. Record the quantities from all the generating units to aid in comparison of the simulation of
dynamic response to that of the on-line generating units.
3. Consider the utilization of more accurate load models, such as the composite load model, and verify
the dynamic generator models to capture the dynamic features of the real system. (In this study, the
PMU waveforms had a slower and smoother recovery than the PSS/E simulation results).
ACKNOWLEDGEMENTS
In addition to the authors of this paper, the analyses presented therein were completed with the assistance
of the following PowerSouth personnel – Jay Farrington and Brandon Horn.
REFERENCES
[1] D. N. Kosterev, C. W. Taylor, and W. A. Mittelstadt, “Model Validation for the August 10, 1996 WSCC
System Outage,” IEEE Transactions on Power Systems, 14(3), 1999.
[2] D. N. Kosterev, “Hydro Turbine-Governor Model Validation in Pacific Northwest,” IEEE Transactions
on Power Systems, 19(2), 2004.
[3] NERC MOD-033-1 Standard “B. Requirements and Measures
Abstract
For years, conventional ultrasonic testing (UT) systems have aided manufactures and utilities in weld
inspection and crack identification, both during fabrication and during in-service maintenance checks.
Ultrasonic technology has advanced over the years but the fundamentals of how sound propagate
through materials have not changed. An important advancement has been the development of Phased
Array UT (PAUT) which enables the inspector to use a probe with multiple elements to focus, steer and
scan with a single transducer, greatly improving the inspection of complex weld geometries.
This paper will discuss the advantages and disadvantages of PAUT and Conventional UT in the utility pole
industry and will show how Phased Array Ultrasonic Testing is a viable technological advancement and
should be included as a tool in the arsenal of the utility pole inspection team.
Conventional ultrasonic flaw detectors have been used for years in the inspection
of materials, fabricated parts, and as an in-service maintenance tool used by
inspectors across many industries including the utility field. As technology
advances, the standards that govern the processes may occasionally lag behind
causing older technologies to be used in lieu of newer systems Phased array
ultrasonic testing (PAUT) is the younger sibling of conventional UT that could
become the method of choice.
Ultrasonic phased array systems can be employed for almost any situation where
conventional ultrasonic flaw detectors have traditionally been used. Weld
inspection and crack detection are the most important applications and these
tests are done across a wide range of industries including aerospace, power
generation, pipeline construction, general manufacturing, and the utility industry.
The benefit of phased array technology over conventional UT comes from its
ability to use multiple elements to steer, focus, and scan beams with a single
transducer assembly. Beam steering, commonly referred to as sectorial scanning,
can be used for mapping components at appropriate angles. This can greatly
simplify the inspection of components with complex geometry. The small
footprint of the transducer and the ability to sweep the beam without moving the
probe also aids in inspecting such components in situations where access for
mechanical scanning is limited. Sectorial scanning is also typically used for weld
inspection. The ability to test welds with multiple angles from a single probe
greatly increases the probability of detection of discontinuities.
The system can be configured to collect data from multiple probes in sequence to
optimize data collection from opposite sides of the weld or multiple positions
from the same side of the weld in one scan (Fig 2).
In contrast to conventional UT, the beam from a phased array probe can be
moved electronically, without moving the probe, and can be swept through a
wide volume of material at high speed. Due to imaging by B, C, S, and L-scans
(Fig.4) the evaluation and documentation of PAUT is easier. The inspection
speed is higher because of fast scanning in combination with electronic beam
steering.
AWS D1.1’s Annex Q gives the engineer that ability to allow PAUT in lieu of
conventional, but a more formal procedure within Clause 6 with
acceptance/rejection criteria would eliminate some of the variations caused by
Annex Q.
Mr. Ehler is a 1998 graduate of Baylor University in Waco, Texas with dual degrees in
Environmental Science and Psychology. After spending several years in the heavy construction
and utility industries, Mr. Ehler was one of the founding employees of ReliaPOLE Inspection
Services Company, a Houston based inspection firm focused on the global pole and lattice tower
industries. In 2015, Mr. Ehler and ReliaPOLE branched out to form Texas NDT Academy. As
TXNDT’s Senior Instructor and General Manager, Mr. Ehler is committed to providing his
students with the ideal curriculum consisting of both the theories and science behind each course
offered, but also their use in real world applications and best practices.
Certifications:
AWS CWI – 13071361
ASNT/ACCP Level III – 229488 UT/MT/PT/VT
NACE CIP I
AGA Certified Inspector
Co-Author:
Mihai Radu, ASNT Level III
About the Presenter: Mr. Mihai Radu
Mr. Mihai Radu obtained a Bachelor in Mechanical Engineering – 5 year University Degree
from the Polytechnic University of Timisoara – Romania. After relocating to United States from
Romania, Mr. Radu began specializing in Quality Control and Non-Destructive Testing. He
obtained ASNT Level III certifications in MT, UT, and PT. Mr. Radu served as the NDT Level
III at Pelco Structural where, in addition to his NDT duties, he acquired working knowledge and
experience with Phased Array Equipment. Currently, Mr. Radu is the NDT/LAB Manager at
TMK-Ipsco in Catoosa, OK.
`
Drones: The Use of Unmanned Aerial Systems for
Surveying for Transmission Route Engineering,
Substation Design and Installation of Facilities
Table of Contents:
Index……………………….....................................................................................................1
Introduction…………………………………………………………………………………..2
FAA Regulations……………………………………………………………………………..3
Airspace……………………………………………………………………………………....6
Aeronautical Charts………………………………………………………………………......7
Types of Drones………………………………………………………………..…………….10
Software……………………………………………………………………………………...12
Conclusions…………………………………………………………………………………..13
Introduction
Drones, were officially termed by the Department of Defense in 2005 as Unmanned Aircraft Systems or
UAS. They are being used for an increasing number of applications world-wide. This paper will review
some of the UAS issues and drone applications of interest to engineers for designing facilities and
infrastructure, conducting inspections, planning improvements and documenting construction.
The combination of using lithium polymer batteries, light-weight materials, wireless technology and GPS
provides the basis for controlled unmanned flight. Advances in camera and LIDAR hardware and
specialized software development have provided a way to capture and process images and translate them
into hard data. This has resulted in an explosion in the UAS industry for the engineering disciplines, as
the use of drones and related technology has proven to be an ideal solution for data acquisition.
Drones are now regulated by the FAA. Revisions to recent regulations make operating drones for
commercial purposes practical. However, the use of drones has created controversy and has raised
concerns. The right to privacy is a hot topic, as is safety and integration into the national airspace.
Previously, photogrammetric mapping was done by using cameras mounted in a standard fixed wing
aircraft flying between 1500 feet and 12,000 feet. The data was gathered, downloaded, analyzed, post-
processed and delivered. The procedure was costly, cumbersome and caused delays. At the present time,
almost anything can be mapped by using UAS/drone based camera imaging and scanning hardware.
Results are immediately available, and real-time conditions can be viewed and documented.
Advances in photogrammetric software allow photo images to generate point clouds, which can be
converted into high quality aerial maps or topographic digital surface terrain models. Advances in
hardware development reducing the size of LIDAR scanners has resulted in a cost-effective way to
conduct 3D scanning and modeling. A variety of industrial sized drones are now available, Cameras are
becoming more sophisticated and powerful and integrated software provides flight planning and
automatic operation. These advances provide the engineer with many choices. This paper will present:
FAA regulations
An overview of present studies and future issues
Flight parameters and restrictions
Key Elements of a UAS Program
Drone types and uses
Camera types and uses
Software development for flight planning and drone operation
Conclusions, references and links
Disclaimer: The contents of this technical paper represent the perceptions, opinions and conclusions of
the author, and may reference commercial, government and non-profit sites. The contents do not reflect
the views and opinions of UTA/TSDOS or Sempco Surveying, Inc. and are not to be considered legal
advice. The information and links may not represent current sites, rules and content. No liability is
expressed, implied or assumed by UTA/TSDOS, Sempco Surveying Inc or this author for any damages
for the use of the contents. The author has not asked for nor obtained any compensation from any person,
company or organization for the use of names or images of products and makes no specific
recommendations. All trade names are trademarked and/or copywritten by their respective companies.
FAA Regulations
In August 2016 the FAA revised the requirements for those who operate a drone for commercial
purposes. They also set guidelines for where you can fly, the maximum elevation for drone flights, the
weight of the UAS system and outline additional requirements for UAS operation.
The FAA rules for operating an unmanned aircraft for commercial purposes are as follows:
Regulatory Basis Title 14 of the Code of Federal Regulation (14 CFR) Part 107
Remote Pilot Certification involves passing a written and proxied test on general airman knowledge, with
an emphasis on section 107 covering the rules for UAS flight. In addition, there are questions on weather,
risk factors, airspace, aeronautical charts and UAV maintenance, care and record keeping.
The UAV team structure is not defined, but there shall be a FAA certified Remote Pilot present, as well as
a Visual Observer at a minimum. Under certain conditions, additional team members would be on hand
for public relation questions, security, safety and for establishing ground control points. Each team
member should be dedicated to that individual role so all tasks are accomplished.
Pre-flight briefings are essential so that all personnel understand responsibilities and have a good working
knowledge of flight parameters, risks, and response options should there be a flight malfunction. A clear
plan of action in the event of an emergency should be part of every briefing, with adequate personnel on
hand for effective response.
FAA Issues
The FAA is currently conducting studies nation-wide on how to integrate drones into the US
Airspace. Everyone from major online distributors to local police departments, motion picture
companies, real estate and news agencies, utility companies, agricultural inspectors, farmers,
photographers, engineers and surveyors operating drones, there will be more interaction between
unmanned aerial vehicles and manned aircraft. In March (2017) a Canadian 767 had made its
final approach in Ft Lauderdale when a crew observed a UAV passing about 10 feet over its left
wing. Two weeks later a West-jet 737 encountered a “rotary UAV” during a departure. (2)
In the last 3 months of 2016, the FAA listed 399 sightings as “hazardous/ or unauthorized
activities.
The FAA’s overall vision is to allow drones to operate into the National Air Space, side-by-side
with manned aircraft, as opposed to prohibiting this technology. This will only occur if there is
cooperation between industry, professional consultants, the private sector as operators, federal
state and local agencies, political subdivisions, and the academic world.
The FAA “UAS Center of Excellence” (COE) has created an organization, the Alliance for
System Safety of UAS through Research Excellence (ASSURE) which is a global think tank of 23
of the world’s leading research institutions working with thirteen universities and over one
hundred leading industry and government partners, to focus on research, education and training
in areas critical to safe and successful integration of drones into the nation’s airspace.
Their mission is to provide the FAA the research they need to quickly, safely and efficiently
integrate unmanned aerial systems into our National Airspace System with minimal changes to
our current system.
The ASSURE vision is to help the Unmanned Aerial System market grow into its multi-billion
dollar market potential by conducting research that quickly, safely and effectively get UAS
flying alongside manned aircraft around the world. (3)
This graphic shows the location of these partners. White Papers are published with results of the
studies and recommendations for FAA review and/or approval. A copy of the ASSURE 2016
Annual report is attached and an electronic link is provided in the reference section. (4)
(5)
Two major issues are BVLOS, Beyond Visual Line of Sight, and flights over people and
structures. Privacy issues occur when flying over parks or neighborhoods. Safety issues surround
flights over highways and anywhere vehicular traffic is present. Data security and management,
adequate maintenance standards, human factors, detection near airports or special events, risk
assessments, and public acceptance and trust are other issues being studied.
FAA Airspace
Airspace is a dynamic issue that needs to be understood to operate a UAV. Airspace is divided into 7
different categories. A through G, with A being that area above 18,000 feet and G being the area from the
ground up to either 700’ or 1200’. The graphic would look like this:
(6)
Generally, a commercial drone can operate from the ground up to 400 feet above ground level. If the
drone is not within 5 miles of an airport, or within a no-fly zone, and if f the FAA’s other restrictions are
followed, a drone may be flown for commercial purposes without contacting anyone. If the drone is
within 5 miles of an airport, the FAA requires the airport to be notified.
There is an exemption process if you need a waiver from the general rules.
The FAA has an app for smartphones called B4UFLY, an “…easy-to-use smartphone app that helps
unmanned aircraft operators determine whether there are any restrictions or requirements in effect at the
location where they want to fly.” (7a)
The FAA also provides a UAS data delivery site which provides shapefiles and maps of the National
Airspace of interest to drone operators. (7b)
Aeronautical Charts
The FAA publishes Sectional Aeronautical Charts which are the primary navigational reference medium
used by pilots of manned and unmanned aerial vehicles. They provide essential information for
navigation to and from airports, controlled space, restricted areas, obstructions and related data. (8)
The airspace in the DFW metroplex is very complex. Part of the FAA Remote Pilot Exam covers reading
these charts and understanding the airports shown, their airspace and flight restrictions, locations of
various hazards and special events status.
This section of the DFW chart shows a number of airports, airport towers and landing strips, Class B, C,
D & E airspace, cell towers, a variety of terrain and manmade obstructions, temporary restrictions,
weather radio frequencies and special flight patterns.
(9a)
The UAS pilot needs to understand the big picture when planning a flight. Locations of helicopter landing
pads and smaller unmanned or non-towered rural airports and knowledge of any unusual events or
emergencies should be part of the planning.
The FAA also works with Arcgis to provide local facility maps where the FAA may authorize UAS
operations without additional safety analysis. They can be used for informational purposes in requesting
waivers.
Applications are widespread. The professional engineer needs to acquire reliable data for proper design.
Using a fixed wing or quadcopter, one can obtain aerial images or LIDAR scans of areas of interest. What
once took a high altitude aerial flight can now be accomplished with a drone flight at a couple of hundred
feet high.
This application is of use in route design as well as in substation design. The concept is relatively simple.
Fly over a site at a certain altitude with a camera or LIDAR scanner, position the lens at a certain angle,
and program the software for overlap of photos, flying in a grid or crisscross pattern. The drone software
makes the actual flight semi-automatic.
With cameras, specialized software will stitch the photos together and convert the pixels to 3D scalable
points, providing a TIN or triangulated irregular network from which a contour map can be generated.
Connect the dots and you have your planimetric side of things. Download into a CAD drawing and you
can inverse between points or scale as necessary. With ground based control points, you will obtain a
georeferenced control diagram which can readily be used for design purposes.
LIDAR scanners, can penetrate ground cover as well as canopies from trees. Post processing software
will generate contours or planimetric drawings.
Utilizing infra-red cameras, drones can be used for inspection by capturing “hot spots” on transmission
lines or in substations. Various components of the electrical systems can be viewed in real time with
video flyovers.
This technology, coupling a drone with a camera set for single photo or video, or with a LIDAR scanner
with a GPS antenna, is readily adaptable for gathering data for substation rebuilds, for additions,
inspections, repair, for generating as-built surveys and for security.
The same technology works well for route design, gathering data for greenfield planning, vegetation
management, inspections, discovering or documenting encroachments or problem areas, emergency
repair, and gathering specific data for re-routing or replacement purposes.
One can conduct circular flights around towers, take shots from a nadir or oblique or 90-degree positions,
shoot video, conduct infra-red inspections for locating hot spots or potential failure areas. Image pattern
or CAD drawing overlay can be used to compare design with as-built conditions.
The drone industry is expanding on all fronts. Numerous companies provide turnkey solutions, and will
construct a drone with matching cameras, lenses and software customized for an individual need. Many
companies will rent drones, and others provide training and logistical support.
Plan of Action-A plan to determine what the needs are, what the solutions are, establish financial costs,
estimate administrative and/or training hours, and establish goals to best achieve positive results.
Leadership-An individual or group to oversee the program and understand the formational needs,
structure, costs and development of the plan, as well as understand the overall regulations. Should also
track issues and revisions to rules at a local, state and federal level.
Structure-Determine how many individuals are needed for goals to be met. A UAS certified remote pilot
is required to be physically on site and part of the team. A visual observer is required whose sole task is
to focus on the drone path and watch for other aircraft, obstacles and anything that might affect the drone
flight. You need a person at the controls who understands the flight path, hardware and software and will
actually operate the flight controller. Field survey crew may be necessary to set ground control.
Operations Manual-Develop a manual to illustrate all the aspects of the drone program.
Safety Program-Part of the Operations Manual, and should include daily briefings as well as post-flight
analysis.
Insurance-Liability at the million-dollar range is common, plus a “hull” rider if the drone is to be covered.
Training Program-This should include FAA Certification as a remote pilot as well as training for all
members of the crew. Each crew member should have specific duties with safety as a priority.
Maintenance Manual-In addition to the manufacturer recommendations, this should include specific
additional steps for keeping all parts of the drone system in exceptional working order. Logs for battery
life and charging dates are necessary as well as an inspection calendar or Pilots Logbook to track flight
hours, intervals for cleaning, maintenance and repair.
Aircraft-Based on the perceived application, budget, and research. Fixed wing or rotary copter, number
and size of batteries, payload, length of flight, transport and storage should be considered.
Camera-Determine the character of the deliverable (photos, video, point cloud, chart), select a camera and
lens. Focal length and number of megapixels are important. Choose a matching gimbal package for
attaching to the drone. The entire package shall be less than 55 pounds.
Software-Determine what is compatible with the camera and what the flight software is included. You can
have multiple software packages operating the drone flight itself, the camera or LIDAR scanner. Flight
software includes establishing a grid pattern to cover the area at a certain elevation, the overlap of photos,
and points of beginning and ending for the flight.
Personnel-Number of field teams, number of crew members on each team. Office postprocessing as well
as administrative needs and ongoing inspection, maintenance and repair personnel.
Marketing-A plan and personnel to provide services to existing clients as well as obtain additional ones.
Types of Drones
There are two types of drones in use today, fixed wing and quadcopter. This paper will focus on
quadcopters as they provide more maneuverability when flying a crisscross or grid pattern. The
quadcopters of interest must have enough stability and payload capability to carry a good camera
or LIDAR scanner. All types listed are trademark proprietary brand Names.
The biggest drone manufacturer is DJI based in China. The top lines are the industrial Matrice
Pro, Dji Phantom and Dji Inspire Series.
Dji Matrice 600 Pro-20 lbs; Dji Phantom 4 Pro-4 lbs, Dji Inspire 2 -4 lbs,
A very good comparison chart is available from the heliguy.com website (10)
Parrott, based in France, has a package available called their Bebop Series. It uses Pix4D software and
provides 3D modeling. (11)
Parrot also owns the SenseFly brand, which offers the eBee fixed wing and albris rotor drone and
software packages. The Albris system has a triple view head with HD video, 38MP still, and thermal
cameras A comparison of the four eBee and single albris systems is available from their website. (12)
In 2015, Yuneec, based in China started a partnership with Intel after a sixty million dollar
investment and announced another partnership that same year with Best Buy. (13)
Their Typhoon Series provides a basis for 3D modeling. The H520 is currently in production. (14)
Yuneec 920 Plus 11 lbs Yuneec H520 ? lbs
As DJI is the biggest seller of drones, they have an extensive gimbal/camera configuration.
Gimbals are the custom connections that mount the camera to the dron e.
Hasselbad, using the Ronin gimbal and Zenmuse, with its own gimbal are compatible with the
Matrice Pro lines. The Ronin MX Gimbal is designed for both the air and ground. (15)
For LIDAR scanning the Riegel Line offers many choices. An IMU (Inertial Measurement Unit) needs to
be attached to the LIDAR Unit and an external GPS antenna, dedicated to the LIDAR package, is also
required. (16)
YellowScan provides a scanner for use with the DJI Matrice 600 among others. (17)
Infrared cameras are used for inspections for locating hot spots. Many companies offer turnkey
solutions. Flir offers several cameras for that purpose. (18)
Software
Pix4D Mapper Pro © – Photogammetry software for professional drone based mapping, purely from
images. Drone images are translated into x,y,z points with highly precise, georeferenced 2D maps and 3D
models as a result. They’re customizable, timely, and complement a wide range of applications and
software. (19)
Advantages:
Intuitive
Ease of Use
Interoperability of File Types
Quick Export of Contours/Surfaces
3D Models
Point Cloud
Flexible Licensing
Video Flythrough
Easy Quality Control/Data Check
A link to the feature list for the Pix4D Mapper is provided in the references (20)
DatuSurvey™ turns drone and camera imagery into accurate, geo-referenced 2D maps and 3D models,
saving the need for expensive and risky field work and expediting deliveries. This 3D mapping software
features a survey-grade accuracy and an intuitive workflow, designed especially for land surveyors and
geospatial professionals. DatuSurvey™ multiple outputs include: orthophoto, projection on 2D planes,
volume calculation, point cloud, 3D texture and mesh model and more. (21)
Advantages:
Easy to Use
Fast Processing
Point Cloud
Epipolar Line projection for quick and accurate point creation
Survey focused workflow
Export Polylines and Points to DXF
Photomodeler © The PhotoModeler Software extracts 3D measurements and models from photographs
taken with an ordinary camera. A cost-effective way for accurate 3D scanning, measurement, surveying
and reality capture. (22)
Sensefly © – bundles with eBee © and albris © drones for seamless flight planning, capturing of images
and generation of data and maps ; final report/deliverable creation in third-party software. (23)
Dronedeploy © - A Software Platform for Drone Mapping; Capture imagery, process maps & 3D
models, and interpret data (24)
Conclusions
Drones are here to stay and will become an important tool in the acquisition of data. Use of UAV systems
allow an engineer to gather data for a variety of uses across a variety of disciplines. To be practical, the
design of a UAV system should be optimized so that each need is addressed. A drone set up for LIDAR
will need an additional GPS antenna permanently attached, in addition to the GPS antennas the drone uses
for location purposes. A drone dedicated for photography could be substantially lighter than one used for
LIDAR data gathering purposes.
A variety of cameras, still and/or infra-red, would provide flexibility in different applications. Purchasing
separate drones with appropriate gimbals or connecting braces for each system would allow versatility in
data collection.
Software is chosen to provide data processing as appropriate for each project and deliverables can match
the engineering or reporting needs.
The UAV technology, the advances in design and the practical applications are constantly changing as are
the regulations and social concerns. Safety is paramount, as is the protection of privacy. With a thorough
understanding of those issues, the knowledge of the elements in a drone program, and a good
understanding of what kind of applications and resources are available, an engineer can make best use of
this technology in theory and practice.
It is now practical to gather data for transmission route engineering, substation design, create as-built
drawings, view encroachments, conduct inspections and track operational issues using UAS technology.
The engineer or design professional has numerous opportunities to acquire this data. Options exist to
either retain a consultant that provides a UAS surveying service or create an in-house service.
There are several national and international organizations devoted to UAS theory, application and
practice, among them are:
AOPA Aircraft Owners and Pilots Association originally established for manned
flight, now has a drone pilot membership option, tracks safety,
regulations and similar issues. (26)
Know Before You Fly Organization between AUVSI and Academy of Model Aeronautics in
partnership with FAA as an educational campaign to enhances safety and
responsibility. (27)
Numerous magazines and periodicals are available providing valuable discussions and information on
current events in the UAS industry; including, Inside Unmanned Systems, Unmanned Systems
Technology, Flight International, Heights (xyHt) and Military Imbedded Systems. Websites are too
numerous to list, but provide additional resources, as do on-line forums, blogs and support groups.
a. B4UFLY faa.gov/uas/where_to_fly/b4ufly/
Introduction
By now most everyone is familiar with the newest technology buzzword that has captivated
aviation enthusiast and industry professionals alike. DRONES! DRONES! DRONES! (The term used in
professional circles is Unmanned Aircraft Systems, or UAS for short). Perhaps you may have already
purchased one for yourself or your children as a present. It is true that the consumer markets have latched
onto UAS on an incredible scale. But the reality is that this amazing technology, though still very much in
its infancy, has already begun to transform a huge number of major industries and will continue to do so
with increasing pace. It seems innovations are announced almost daily that make drones smaller, smarter
and more valuable to those of us in the dull, dirty and dangerous areas of Utility O&M. But with so many
new ideas and opportunities lingering just around the corner, it is important to understand where the
technology is today and how it can best be used currently by utility industry professionals like you.
The Airline to Holly transmission line inspection and re-conductor project is a perfect example of
how drone based aerial inspection techniques can help improve the operational capabilities of a line
owner. In this case, an older line that has been in service for a number of years on the gulf coast is due
for installation of a second circuit. A conditional assessment of the structures is essential before
construction can begin. Corrosion, cracked or damaged welds and other structural deficiencies can be
identified and made known to the project engineers prior to crews arriving on site to begin work. This
operational insight can add tremendous value to the team responsible for maintaining the line. But what
is the best method for performing the assessment when the line is situated along a busy thoroughfare
with several structures located in someone’s back yard? In addition to the “urban” nature of the line
location, there is also an International Airport, Naval Air Station, and two private airstrips all in close
proximity to the project area. This case study presentation will outline the various challenges and
solutions related to conducting drone operations in this type of congested environment, including
obtaining FAA authorization. It will also address the critical needs experienced and expressed by the line
operator and the key areas where they found the most value in deploying drone technology on their
system.
The Airline to Holly Transmission Line is a 2.2 mile double circuit (138kv/69kv) line segment that
was built in 1969. The 138kv circuit is still in service but the 69kv circuit was recently retired and AEP is in
the process of analyzing the feasibility of installing a new 138kv circuit on the existing structures to replace
Figure 1
Figure 2
With many options available to AEP to perform the inspections, there were several important
factors that made the use of drones the optimal choice for this project. The first reason is that parts of
the line segment were located in boggy wetland areas where access via bucket truck was not feasible.
With a UAS and a light footprint vehicle, access to the structures was increased significantly. Likewise,
other structures were located in property owners’ back yards and again access was limited. The fact that
UAS can be remotely operated and provide access to “hard to reach” areas is a major value proposition
for the technology.
The second and third factors in the decision to use UAS were the cost and resulting inspection
deliverable. For this type of targeted inspection, UAS operations tend to be much more cost effective that
helicopters, cranes and man lifts. And in addition to the cost of inspection, the deliverable is of higher
quality than what you would be able to capture from a walking inspection or aerial survey from a manned
aircraft. With a drone, the operator can position the craft only feet away from the tower, making it
possible to detect hairline cracks or other very small but potentially catastrophic defect. The Falcon 8 UAS
shown in Figure 3, which was used for this inspection project, caries a high resolution digital payload (Sony
A6000 DSLR) which is optimal for high detail inspections.
Figure 3
There are many aspects of line maintenance that make the operation challenging. Working at
heights, working in an around high voltage equipment, environmental conditions, and in some cases
regulatory requirements an all play a role. Each have some type of impact on the related work being
performed. As previously mentioned, UAS technology can help mitigate some of those risk factors such
as performing inspections remotely from the ground, keeping the technician safely away from hazardous
or potentially life threatening situations. It can also improve the quality of inspection data and drive up
the probability of defect detection so that small but critical deficiencies are not missed by the inspector.
And finally, it allows inspection work to be performed on energized lines which eliminates the need to
take costly outages. Like anything else, however, there are tradeoffs. UAS operations do present
challenges of their own such as onerous regulatory guidelines to follow, airspace challenges, and certain
risks inherent with flying in urban and congested areas.
The use of UAS for commercial purposes is heavily regulated by the Federal Aviation Agency (FAA).
In August of 2016 the FAA released what is known as “Part 107” of the Code of Federal Regulations (CFR).
In it there are very strict guidelines on how and where you can legally operate these types of aircraft.
Some of those include flying within visual line of site (VLOS), flying during daylight hours, flying under a
maximum altitude, restricting flights over people and only flying in class G or uncontrolled airspace.
The operator must always maintain visual line of site of the aircraft during flight. This means that
the craft must remain within a certain distance of the operator where it can be seen by the naked eye.
Aids such as scopes, binoculars or cameras are not permitted to satisfy the requirement. For small aircraft
such as the Falcon 8, our operators typically maintain VLOS out to approximately 1500 feet. It also means
that you cannot fly behind structures that would obstruct line of site from either the pilot or the observer.
In certain urban areas where ground access for the operator is limited this can present challenges to the
inspection team.
Additionally, all flights must be performed during daylight hours. This condition has less of an
impact since visual inspection would become more difficult in low light situations, however it does prevent
certain IR scans that would need to be performed at night when solar reflection and loading (factors that
negatively impact IR readings) are mitigated. The Airline-Holly inspection did not utilize IR imaging so this
requirement did not impact the project but should be considered on any IR inspection that would benefit
from night time flying.
The craft must also be kept below 400 feet above ground level (AGL) in class G airspace. A caveat
does exist for tower inspection that allows for UAS flights above 400 feet AGL if the craft remains within
400 feet vertically and laterally from the tower in question. The Airline-Holly poles being inspected were
all under 100 ft AGL so there was never any concern about satisfying this requirement. However, the pole
height did play a significant factor in the airspace waiver that will be discussed later.
Figure 4
To prevent pedestrians from encroaching the operational area, perimeters for each flight are also
established using cones, flags, and signs to clearly cordon off the area. It is important to remember that
despite the overt warning messages presented by these perimeters, often pedestrians will simply ignore
them and procedure through. This is especially true if the perimeter blocks a bike path or sidewalk. In
areas such as this, it is essential to have a third crew member present for “crowd control” during flights.
It is not safe or even functionally possible for the pilot or inspector to keep the perimeter secure while
they are flying the aircraft and performing the inspection. Typically, as was the case for this project, flight
times are only between five to seven minutes of duration. Any pedestrian can simply wait for the craft to
land or find another route around the perimeter.
Another feature of the Airline – Holly line that required consideration was the location of poles in
and around the property of local businesses and home owners. In rural areas this is not a major issue but
Figure 5
Figure 6
Communicating operations and maintenance activity to affected land owners and the public in
general is always a good practice. Keeping people informed is a proactive step in minimizing surprise or
concern when it comes time to perform the work. This is especially true when the work involves new
technology that can seem menacing to certain people who are not well informed of the intended
application. You can imagine the potential surprise of looking out your window to see a drone flying
around you back yard! If you are not aware of the work being performed and why, then you might become
agitated or alarmed. That is why on projects such as this, a well-coordinated notice to land owners that
outlines the type of work being done and by whom, as well as the appropriate licenses and permits
applicable will go a long way to ensuring a smooth project. For the Airline to Holly inspection, AEP agents
sent out notifications to all affected land owners prior to the scheduled work date which identified the
Figure 7
Because the project area fell squarely inside Class D airspace, in order to legally inspect the
transmission line we had to secure a waiver that granted the operator authorization to fly the UAS in
controlled airspace. There are two primary ways to receive this type of authorization from the FAA. The
first is to apply for what is known as an emergency certificate of authorization (ECOA). This will only be
granted for projects where loss of life is eminent or a public emergency has been declared. Any drone
operations that would take place following a major storm event would likely fall under this category. For
non-emergency situations, such as routine inspection, the operator is required to submit a request of
waiver to the FAA UAS office. In both cases, the application must show a very detailed map of the flight
area and the proposed altitude, time, and nature of the flight. Instructions on how to properly fill out this
Inspection Results
Once the FAA granted the airspace authorization required to fly the UAS, the field work was
scheduled for April 6th of 2017. The weather that day was well within operational limits of the UAS with
winds holding steady around 12-15mph and partly cloudy skies. For reference, the maximum wind
threshold for the Falcon 8 aircraft is around 30mph. A total of 36 monopole structures where inspected
over approximately 9 hours of inspection work. The actual flight time for the UAS totaled 4.2 hours. The
inspection was focused on all of the structural elements of the towers including the foundation, slip
splices, seam welds, arm vangs, vang welds, davit arms, and arm shaft to bracket welds.
The findings from the inspection were consistent with what we would expect on a line of this age
and in this environment. There was a noticeable level of surface corrosion on the structure itself and the
hardware. The davit arms, which had already been identified as needing replacement, showed signs of
major corrosion at the arm shaft to bracket weld. In fact, some of the corrosion was so severe that certain
arms likely would not last until the proposed construction schedule and needed to be replaced
immediately for safety reason. On the surface, the pole structure and arm vangs appeared to be in decent
condition. There was no evidence of severe corrosion and/or weld cracks that would cause the structure
to fail at the vangs. However, the internal condition of the structures remains unknown. There were
several poles with missing pole caps which would allow further moisture ingress. The hand hole ports at
the base of the pole would likely need to be opened to ensure that there was not any standing or trapped
water at the base of the pole which would cause significant corrosion and wall thickness loss in the base
section at ground line. Figures 8-11 show the varying levels of corrosion found on the arms as well as the
general condition of the arm vangs, slip splices and ground line embedment.
Figure 9
Figure 11
The final outcome of the UAS inspection for Airline to Holly has yet to be determined. The
major takeaways from the data that was collected were focused primarily on the davit arms. While it
was already decided that the arms would need to be replaced before the reconductor work could begin,
the aerial inspection confirmed that several of the arms that were currently in service needed to be
replaced immediately due to sever corrosion. Visually, the other structural components of the poles
showed no defects. There were no visible cracks present in any of the arm bent brackets, vangs, vang
welds, or slip splices. The foundations appeared to be in adequate condition as well. However, the
internal condition of the structures and potential wall thickness loss due to corrosion cannot be
determined without a more through inspection (either straight beam UT at ground line or a visual
inspection through the hand hole ports). AEP is still in the process of analyzing the data to determine
whether the monopole structures will need to be replaced in addition to the arms, or if their life can be
extended further to support the new 138kv double circuit line.
It is clear, however, that UAS as an inspection tool added significant value to the line owner in
collecting the necessary conditional and structural assessment data on the Airline to Holly monopoles.
Despite the project being in a crowded, urban environment with controlled airspace near several major
airports, the inspections were completed safely and legally with the appropriate FAA approvals. The
costs of the inspection where minimized by eliminating the need for cranes, outages and climbing.
Overall, the Airline to Holly project proved that UAS can be deployed effectively in urban environments
and ultimately add significant value to the line owner and their operations team.
1 1 1
0.5 1 0 1.5 2
Top of Pier Rotation Limits [Deg]
Weighted Average = 1.6 deg (Total 10 Respondents)
Figure 3: DFI Survey Results on Pier Rotation 1
Number of Respondents using
2
particular deflection limit
2
1
2
1
1 1 1
1 1
0.0 0.5 1.0 1.5 2.0
3.0 2.5 3.5 4.0 4.5 5.0 5.5 6.0
Deflection Limits [in]
Weighted Average = 2.55 in (Total 14 Respondents)
This paper will focus on the foundation performance criteria of dead-end structures since they exhibit the
highest consequence of large foundation rotations and tip defections. Some very good papers have been
written regarding different foundation performance criteria for different structure types. One that this
author highly recommends is “Recommendations for Steel Pole and Drilled Shaft Deflection Criteria Based
on a Parametric Study of a 138kV Double Circuit Transmission Line”. 3
LOAD CASES ANALYZED TO MEET PERFORMANCE CRITERIA
Table 1: DFI Survey Results on Load Type for Checking Performance Criteria 1
Figure 6: MFAD - Components of Lateral Resistance4 Figure 7: MFAD – Pier Deflection vs. Depth
Lpile
Figure 8: Lpile - Components of Lateral Resistance Figure 9: Lpile - Pier Deflection vs. Depth
As can be seen in Figure 8, Lpile assumes that only p-y curves provide lateral resistance. For piers with
L/D ratios greater than 10, p-y curves make up the majority of the lateral capacity. Another Lpile
assumption is that the pier does not exhibit rigid body rotation. Lpile assumes that the concrete will flex
and crack when the applied moment exceeds the cracking moment capacity of the pier, as illustrated in
Figure 8 and Figure 9. This assumption creates a far more complex design model that requires numerous
iterations to converge upon a solution. This assumption results in pier rotation varying with pier depth.
This assumption can result in larger pier tip deflections and rotations than MFAD’s rigid body rotation
design assumption.
RIGID BODY ROTATION – FULL-SCALE TEST RESULTS
One paper that discusses in detail the Rigid Body
Rotation full scale test results is the Evaluation of
Performance Criteria for Short Laterally Loaded
Drilled Shafts.2 Much of the next paragraph refers
to the data presented in this paper. In the early
1980’s EPRI funded research to study laterally
loaded shaft design in the utility industry (Davison,
1982). This research program included field tests
to develop a database of high-quality full-scale
foundation tests in various soil conditions. 14
drilled piers of varying sizes were tested in a variety
of soil conditions. The L/D ratios for these 14 piers
ranged from 2.7 to 4.4.
Figure 10: Typical Shaft Deflected Shape
From this full scale test data, the assumption appears valid that drilled piers with relatively small L/D ratios
display rigid body rotation under lateral loading.
Figure 11: Total Deflection and Rotation Figure 12: Non-Recoverable Deflection and Rotation
EXAMPLE OF UNBALANCED PERFORMANCE CRITERIA
The DFI survey results, (See Figure 3 and Figure 4) show that limits on foundation tip deflection were more
common than limits on foundation rotation. Also, in the experience of the author, many utilities’
performance criteria specify relatively stringent limits on pier deflections as compared to the liberal limits
for pier rotation.
For example, two utilities familiar to this author specify a maximum tip deflection limit of 3 inches and a
maximum tip rotation of 1.5 degrees. It is worth noting that these utilities do not specify any limits on non-
recoverable rotation or deflection. It is also worth noting that total tip deflection limit of 3 inches and a
maximum tip rotation of 1.5 degrees is fairly close to the weighted average values for tip deflection and
rotation from the DFI survey. The DFI survey’s weighted average for tip deflection limit is 2.55 inches
and the weighted average tip rotation limit is 1.6 degrees. MFAD analysis assumes that a loaded pier
experiences rigid body rotation, as full-scale tests have shown. Please refer to Figure 10.
The point of rotation is very much dependent on the soil profile and the lateral capacity of the different
layers of the soil. The point of rotation usually occurs at a depth between ½ and ¾ of the embedment length
of the pier. For this example, it will be assumed that the point of rotation occurs at 2/3 of the embedment
length of the pier. If a typical pier has an 8ft diameter, the 3 inch tip deflection limit can be converted to a
percentage of the 8ft pier diameter.
Table 3: L/D Ratios for 3in Tip Deflection and 1.5deg Rotation that Yield a Balanced Performance Criteria
In the experience of this author, the majority of foundations typically have L/D ratios between 3 and 5. The
required L/D ratios are highly dependent on the size of the transmission line structure that the foundation
is supporting and the lateral soil capacities which surround the pier. If the soil conditions are extremely
good or extremely poor, then the pier could have an L/D ratio outside of the 3 to 5 range but, in the
experience of this author, this is rare.
In Table 3 only the 4 foot diameter pier has an L/D ratio within this “3-5” range where either the rotation
or the deflection limit could be exceeded. Also shown in the bottom row of Table 3 is the pier length in
which a “balanced” performance criteria could occur. As can be seen, the length of the pier remains
constant at 14’-4” for all pier diameters. This is due to the single 3 inch tip deflection limit for all sizes of
pier diameter. It is highly unlikely to have an 8ft or greater diameter pier with only 14’-4” of pier length.
Table 3 shows that the performance criteria of 3 inch tip deflection and 1.5 degrees rotation has a relatively
stringent deflection limit and a relatively liberal rotation limit. This example of stringent tip deflection
limits and liberal rotation limits is not uncommon.
Figure 4 shows from the DFI survey the respondent’s weighted average for foundation tip deflection is 2.55
inches and Figure 3 shows that the weighted average for foundation rotation is 1.6 degrees. These small
deflection limits often result in pier tip deflection controlling the design of drilled piers.
Figure 14 displays the pole displacement components from both foundation tip deflection and foundation
rotation. Four locations along the pole were selected to calculate the additional pole displacement from
foundation movement. The four locations represent the shield wire attachment point at the top of the pole
and the top, middle and bottom conductor attachment points on the pole. As examined above, this next
example will also analyze an 8-foot diameter pier. The limit for pier rotation is 1 degree and the tip
deflection is 4 percent of the pier diameter (4%*96 inches= 3.84 inches).
It is important to note that the 1 degree rotation limit
is significantly more stringent than the 1.6 degrees
weighted average limit shown in Figure 3 and the
3.84 inch deflection limit is far more liberal than the
2.55 inch weighted average limit shown in Figure
4. Using the same process as was stated in Figure
13 and Table 3 we can determine if this
performance criterion will be balanced in regard to
tip deflections and rotations. Once again, assuming
the point of rotation occurs at 2/3 of the embedment
depth then:
4%
= = 3.44
2
sin 1.0 $%
3
Figure 14: Additional Pole Displacement due to Foundation Tip Equation 2: L/D Ratio from a 4%*(D) Tip
Deflection and Rotation Deflection and 1deg Rotation Limit
Table 4: Foundation Tip Deflection & Foundation Rotation's effect on Displacement of a Pole.
REFERENCES
1
Electrical Transmission and Substation Structures Conference SEI & ASCE (2015), Study of Electric
Transmission Line Deep Foundation Design. P.M. Kandaris, P.E., and Davidow, P.E.
2
GeoCongress 2012 ASCE 2012, Evaluation of Performance Criteria for Short Laterally Loaded Drilled
Shafts. P.M. Kandaris, P.E., A.M. DiGioia, Jr. Ph.D., P.E., and Z. J. Heim, P.E.
3
Electrical Transmission and Substation Structures Conference SEI & ASCE (2015), Recommendations for
Steel Pole and Drilled Shaft Deflection Criteria Based on a Parametric Study of a 138kV Double Circuit
Transmission Line. Adam G. Bowland, Ph.D., P.E.; Paul G. Class, P.E.; A.M. DiGioia, Jr. Ph.D., P.E
4
IEEE Std 691-2001 – IEEE Guide for Transmission Structure Foundation Design and Testing,
Transmission and Distribution Committee of the IEEE Power Engineering Society, December 6th, 2000.
Introduction
Commonwealth Edison (ComEd) is large electric utility in Illinois, serving approximately 3.8 million
people, and encompassing approximately 12,000 square miles service area. This utility initiated a
project to update and improve the fault location process. The existing process relied on two or more
separate groups to repair any faulted cables.
First, the Distribution Testing (DT) group would be dispatched to locate the fault. This typically 2 person
crew would identify the faulted phase, identify the faulted section of cable, and finally pinpoint the
location of the fault on the isolated section of cable. Once the section was identified, this crew would
depart, and the next group would take over.
One of three groups would take over from the DT group. The Distribution Overhead (DO), Distribution
Underground (DU), or a contractor group would be assigned to repair the faulted cable. This work was
to be done at an unspecified future date, except in emergency conditions. The problems encountered
with this system of working are described below.
The new method of working will be to dispatch a single crew that is responsible to identify the faulted
phase, identify the faulted section, pinpoint the fault, repair the fault, and ultimately return the cable to
service. This method of working has some initial costs involved, but also substantial gains. These will be
described below.
Existing System
Figure 1 shows a block diagram of the existing process. This complicated process involved multiple
different crews and required handing off responsibility between them. The general workflow of the
existing system is described in Figure 1. In this system, the DT crew was a highly skilled group of fault
locators, consisting of 48 people equipped with truck mount fault location equipment ranging from 1600
to 3200 Joules, and up to 36 kV; sufficient to locate almost any fault.
In this system, multiple crews were required to compete the work. First, Distribution Testing (DT) would
locate faults in the URD system with a 2-person crew. Once the fault was located, they would depart
the scene and turn responsibility for repair over to the Distribution Overhead (DO), Distribution
Underground (DU), or outside contractor crews. One of these crews would be dispatched to the fault
location at a later date. If it was an emergency, or important circuit, they would be dispatched quickly.
But for less critical faults, there may be some delay between DT locating the fault and a crew arriving to
repair the fault.
Occasionally, this delay would result in the repair crew being unable to locate the fault and they would
require DT to return to the fault and locate again. This duplicated work that was already done, and
prevented the DT crews from finding other, newer faults, and generally contributed to inefficiencies in
the process.
If the crew was able to successfully repair the fault, sometimes they would have a questionable result on
the pre-energization testing. In this case, the DT crew would again need to return to the scene to
perform a more thorough test and determine if the cable can be returned to service. This would again
increase inefficiency as the DT crew would need to return to a scene they already completed their work
at.
The five highlighted segments in Figure 1 were eliminated in the new system. This reduced the amount
of crews to one, and streamlined the entire process.
New System
Figure 2 shows a block diagram of the new process adopted by ComEd. Under this system, one
department has responsibility for the entire outage: from fault to fix. The new system would involve
over 1000 employees in the construction group. Prior to this project, this larger group had no fault
locating equipment and no previous fault location experience. One of the design considerations for this
project was to consider the training requirements of the new fault location equipment. Equipment that
was too difficult to learn and operate would not be effective in this system.
The advantages this system has over the previous system are primarily due to the idea that only one
crew has responsibility for the fault from beginning to end. There would be no need to return to the
fault location because the repair crew was unable to locate the fault. There is no transfer of
responsibility from one group to another, simplifying operations and consolidating responsibility in one
group.
The challenge of implementing this system is due to the large number of personnel that are involved.
The new system needs to accommodate personnel with various skill levels, comfort levels, and
knowledge about fault location. In order to minimize this effect, it is necessary to consider the ease of
operation of the fault location equipment, and choose only those with a simple and easy to learn
interface.
Transition
The existing system had two distinct fault location groups: City and Suburban. The city groups were
equipped with 3200 Joule fault locators, and would locate mostly on main stem circuits of approximately
20,000 feet average. The suburban crews were equipped with 1600 Joule fault locators, and would
locate mostly on URD and main stem circuit of approximately 5,000 feet. Both crews utilized truck
mounted fault location equipment.
ComEd has 21 reporting centers where line crews reported day to day. After analyzing the activity at
each reporting center, it was determined that 3 to 5 fault location systems would be needed at each
location to keep up with the amount of faults in the system. All together this would mean about 40 new
fault location systems to fully outfit the construction crews.
If 40 new systems of the same specification as the existing equipment were purchased, it would cost
approximately $4 million: 40 units at a cost of about $100,000 per unit. This is just the purchase cost of
the equipment, training costs, and additional vehicles would be in addition to this cost. A better, more
cost efficient system was needed.
By equipping all the crews with the portable systems, the cost of the transition was reduced to
approximately $600,000 for 40 units, and no additional vehicles would need to be purchased. Training
costs are not included in this figure, but due to the ease of use, the training time and expenses can be
kept to a minimum.
There is always a concern when using such small and portable units that the unit will not have enough
voltage or Joules to locate the fault, and this will happen from time to time. The strategy to mitigate
this risk is to deploy progressively more powerful but fewer units strategically throughout the system
(See “Satellite System” section below).
Work within ComEd continues with this transition. Training approximately 1000 inexperienced
personnel was completed in about 3 months, with the help of the vendor. Classes were held both in the
classroom and in the field to give theoretical and practical knowledge to the crews. Currently, the
construction crews are locating approximately 90% of all faults without the assistance of the
experienced Distribution Testing group.
Future work will address the necessity and viability of large portable units, and large truck mounted
units within the overall distribution system.
Benefits/Improvements
There have been many benefits of moving to the new system. The single largest improvement was in
reducing the amount of handoffs, which resulted in faults being repaired much more quickly. Wasted
time due to redoing previously completed work and travel has been reduced as well, as crews do not
need to revisit a fault multiple times. By having fault location equipment readily available, crews can
arrive at the scene of a fault quickly, and well prepared. This is expected to reduce the overall outage
durations, and work continues in tracking this information.
The disadvantages of these small system is the limited voltage and Joules available. However, these
units also have some particular strengths over the larger equipment. One of the biggest advantages is
that there is no need to dedicate a specialized vehicle to fault location activates. These small units can
fit into a more general purpose vehicle. Also, there is little need for long cable reels, as is common with
truck mount systems, as the small units can simply be placed where they are needed. Finally, the small
units typically cost considerably less than large units, and so more units can be purchased and spread
around the system, leading to decreased response times and high availability of equipment.
Satellite System
This is an example of the start of a “satellite” system for deploying fault location equipment. No one
piece of equipment is best in every situation. The small, light, and portable equipment is ideal as a first
line of defense. This equipment can be deployed for relatively little money, and can be widely and
readily available to the crews that would most need it. For example, putting a portable unit on each
underground truck ensures that no matter which crew is dispatched, they can begin fault location
activities. It also means that a crew can be dispatched directly to a fault, without returning to the
service center to pick up fault location equipment. However, the drawback of these units is their
capabilities. In order to maintain the small size, they often sacrifice maximum voltage, or settle for a
lower amount of Joules. Some faults (On 15 kV systems, this may be less than 1 in 10 faults) will simply
require “more” than the portable equipment is able to provide. That could be more voltage, more
Joules, or more pre-location modes. So, while the small portable equipment will find the vast majority
of faults, especially on 15 kV class URD systems, it is important to have a back-up plan in place for the
more difficult faults.
This back-up system would be large portable units, or a small truck mount unit, depending on the
specific requirements of the end users. There might be only one of these types of equipment in a given
region, and they will typically be used far less often. They are most likely positioned at a convenient
service center, always fully charged and ready to go. These would be far less common in the system, but
would still be easily accessible. When the equipment is needed, someone from the crew would need to
return to the service center, collect the equipment, then return to the fault location, or have the
equipment brought out to them. This may take some hours, depending on the geography of the
distribution system, but it is much more capable than the rapid response, small units.
However, there will still be an occasional fault that even these large portable units will be unable to
locate. For example, a splice failure on a long 35 kV cable would likely be too much for even the larger
portable units. Similarly, the smaller units would be insufficient for faults in an underground network
system due to the large amount of cable-feet in such a system. For these conditions, it is necessary to
have a large, truck mount unit that is capable of finding almost any fault. There would only need to be a
small handful of these in a system, and they will take some time to deploy to any given fault, but once
they get there, it is very likely that they will be able to locate the fault. Strategic positioning of these
type of units can help reduce the deployment time. For example, a system like this is most likely to be
needed for a network fault, so it would be positioned in the closest service center to a downtown
network. It can then respond to the majority of its primary work quickly, but is also available if needed
for difficult fault location activates at more remote locations.
This “satellite” system then can be imagined as having the large truck mount unit in the center, with a
few of the large portable units “orbiting” around it, in various service centers. Each of these large
portable units would then have a few small portable units “in orbit” on the various trucks dispatched
from the service center. This system allows for a rapid response time to most faults, especially in the 15
kV system, but also accounts for the faults that the small units will not be able to locate, and positions
progressively larger units throughout the system (See Figure 3).
Conclusion
By contacting vendors and other utilities, and carefully considering the available options, ComEd was
able to cleverly transition their fault location crew from a small, but highly skilled group to a more
general, larger group. This was accomplished for less than 25% of the first-pass expected costs, while
also improving response time, decreasing time from fault to repair, and achieving the training objectives
easily.
ComEd is also positioned to deploy a full Satellite System, taking advantage of the relative strengths and
weaknesses of each type of fault location equipment to quickly and reliably find faults in the distribution
system, while maintaining a reasonable cost for deploying these systems and achieving a high
availability of equipment.
Abstract
When overhead transmission lines cross regions where either natural or industrial contamination prevail,
airborne dust or conductive contaminants can impact the electrical performance of insulator strings. The
global term « pollution » describing this phenomenon is very often assimilated to anything that comes on the
surface of an insulator, but such « pollution » should be defined precisely to allow for the most appropriate
counter measure selection. Among the solutions adopted by utilities the increase of leakage distance is the
most common method. Likewise, the use of a hydrophobic material such as silicone rubber has proven to be
an effective tool.
We will explain the key concepts and definitions related to insulator pollution problems and solutions
describing how they work and what their limitations can be.
When dealing with pollution problems, the first set of actions is to know what type of pollution the line is
facing. Terms such as ESDD and NSDD are defined. Sampling techniques are described as well as ranking and
classification charts established to determine the severity of the environment.
Once these concepts are clarified, it is possible to engage into the next phase by selecting the most
appropriate insulator type and design. The mechanism of dry band arcing and dynamics of arcing will be
examined showing that leakage distance is not the only parameter to take into consideration. Shape matters
greatly and we will show how different shapes with the same leakage distance can generate different
behaviors and withstand properties.
The use of silicone rubber on the surface of an insulator has proven to work well preventing flashovers from
occurring in a variety of contaminated environments. To better understand how hydrophobicity can help
mitigating pollution problems but also understand how such material can suffer and age prematurely in harsh
environments we will describe the physico-chemical processes at work.
Finally, and while polymers have shown their own limits in terms of resilience (especially in very challenging
environment) alternative solutions such as silicone coated toughened glass insulators will be discussed
including field experience for more than 20 years in extreme conditions with outstanding performance.
Pollution is a very generic term when it comes to outdoor insulation contamination. A general approach
defines pollution like anything that comes and stays on the surface of an insulator which modifies the surface
conductivity, therefore increase the risk of an electrical arc bridging the insulator itself. The deposits are
usually airborne dust, sand, industrial airborne particles, soluble elements such a salt from the sea when a
line is near a coast…
The challenge for line design engineers and maintenance crews is to know how such electric flashovers can
be avoided by either using insulators catching less contaminants, more resistant to arcing or designed and
made with materials which will reduce the ability of an external arc to take place.
When dealing with pollution problems experts use several definitions which are listed hereafter:
Creepage distance: (sometimes the word “leakage distance” is being used). It is the distance an arc needs to
cover from one metal end fitting to the next following the surface of an insulator (figure 1).
ESDD is the acronym for Equivalent Salt Deposit Density. It represents the quantified value of the
contaminants which dissolve in water like salt. When salts dissolve in water the conductivity of the
humid/wet film on the surface of the insulator increases, and subsequently the risk of having a flashover
increases as well. The unit of measure is mg/cm²
NSDD is the acronym for Non-Soluble Deposit Density. It corresponds to the amount of deposit on the surface
of the insulator which does not dissolve in water and remains inert on the surface. Sand for example is a
typical component of NSDD. The non-soluble component of the pollution layer will work like a sponge
catching more humidity which in turn will facilitate the conductivity on the surface of the insulator
Leakage current: is the electric current flowing along the surface of an insulator along the leakage distance.
Each insulator type has a limit above which the leakage current will lead to arcing and possibly a complete
bypass of the insulator (this final stage of arcing is then called flashover).
Specific leakage distance: the ratio between the leakage distance of an insulator (or string of insulators) and
the phase to ground voltage. The unit is inch/ KV and referred to as USCD (Unified Specific Creepage Distance)
Dry band arcing: is a dynamic process by which an arc once initiated will dry the area where it takes place. By
drying a segment of the insulator, the arc will progressively have more and more difficulties to be sustained,
will self-extinguish until the surface is wet again, leading to currents which will generate new arcs, new dry
spots….
The evaluation of the pollution level in any given area on an insulator can be made by measuring both ESDD
and NSDD. The procedure consists in washing the surface of the insulator with deionized water which
conductivity is measured prior to washing. After washing, the conductivity is measured again, and tabulated
for conversion in equivalent salt weight. This number is then divided by the surface which was washed. Once
this is done, the washing water is filtered on a paper filter which needs to be dried in an over. The weight of
solid deposit is divided by the surface of the insulator to give the final NSDD in mg/cm². (figure 2)
Once ESDD and NSDD are defined, it is possible to establish the pollution class as per IEC 60815 (figure 3).
For each class of pollution IEC gives an average USCD as a reference base value for selecting a leakage distance
in the environment under consideration (figure 4).
Insulators exist with different shapes and different leakage distances. Depending upon the local environment
one shape might be a best fit than another one. Some shapes will catch and retain more dust while others
are more adapted to fight arcing activity in moist, humid or foggy conditions. Figure 5 shows a selection
method based on shapes.
Figure 5: recommended shapes for various pollution environments (shapes as per SEDIVER catalog)
Another important aspect in pollution mitigation is the fact that shape will dictate the dynamics by which the
leakage current and the dry band arcing will behave. Figure 6 shows an example where insulators with the
same leakage distance have very different flashover values and very different ultimate leakage currents
threshold values. It is easy to understand that it is not possible to set a maximum reference leakage current
value without defining the type and shape of insulator under consideration.
Silicone is often described as an ideal material to reduce risks of flashover. This is the consequence of the
ability of this material to be water repellant. Hydrophobicity is the generic term used to describe this
property. A material is considered as hydrophobic if the contact angle on the surface of the material exceeds
90° (figure 7). A material with this property will remain dry, shed water droplets and therefore will keep the
leakage current at low values, preventing flashovers from happening. Silicone compounds have also the
ability to transfer this property to any contamination on its surface. This mechanism called “hydrophobicity
transfer” is the result of low weight molecules (LMW) inherently part of the silicone compound migrating to
the surface and embedding the pollutants on the surface. The surface can remain water repellant even with
surface contamination.
Figure 7: Hydrophobicity
While these properties seem extremely attractive, silicone shows also some weaknesses. Among those is the
fact that silicone is relatively sensitive to dry band arcing, leading to erosion and possible cracking of the
material (figure 8). Electric arcing under the form of dry band arcing or corona (result of the breakdown of
air under intense electric field) generates ozone which in turn reacts with nitrogen in air to form nitric acid.
To reduce such effects silicone can be designed with additives such as fillers (like Alumina Tri Hydrate, ATH)
slowing down the degradation process.
While in the past silicone polymer insulators were expected to last as long as porcelain or toughened glass,
reality from the field has proven otherwise with life expectancy down to 15 to 20 years, sometimes more,
sometimes less. Additionally, counter measures such as grading rings to reduce the electric field on the
silicone compound have been implemented at voltage levels much lower than initially thought (as low as 115
KV or even 69KV in some cases) reducing arcing distance and clearances.
Field experience has shown that extreme pollution can be handled with silicone, but at the same time it has
been established that more the contamination is severe the faster the degradation and risk of failure of
polymers. On the other hand, the resilience of toughened glass is known to be a major asset for overhead
lines stability. Combining both has been a concept introduced more than 20 years ago, with great success.
Silicone coatings work best when applied in an industrial controlled environment. In some cases, the
application is made in the field but thickness of the coating as well as adherence remain out of control. In a
factory environment both parameters are carefully controlled. Thicknesses in the range of 10 to 15 mils.
Adherence can be checked with EN ISO 2409 or through a water boiling test as described in IEEE 1523.
The performance under pollution has been established in laboratory tests either with salt fog conditions or
with solid deposits having an ESDD/NSDD under clean fog conditions (figure 9). It was shown during these
tests that toughened glass insulators work extremely well even when the coating is applied only on one side,
underneath the glass bell (figure 10). This is something some utilities might consider for handling purpose
(less damage to the coating when mishandled). Note however that the performance of the coated insulator
is not degraded even when small tears or damage occur to the coated insulators.
Field performance confirms these results with more than a million units used only in Italy along the coasts
removing the need for washing for more than 12 years. Likewise, in the Middle East, silicone coated glass
insulators have been successfully in service for about 20 years. Samples have been removed from the line for
a performance check and the results show that these units outperform after 20 years new non-coated
insulators tested with artificial pollution deposits equivalent to those measured on the line (figure 11). In the
USA the same trend exist, and large utilities are now using either fully coated or half coated units in
replacement of polymer insulators.
Figure 10: silicone coated toughened glass insulator and half coated version
Figure 11: Pollution performance of coated glass insulators after 20 years in service
1
Engineer, Lower Colorado River Authority, Austin Texas; E-mail: Jessica.mclarty@lcra.org
2
Project Engineer, HDR Inc., Ann Arbor, Michigan; E-mail: james.childs@hdrinc.com
3
Assistant General Manager, Valmont Newmark, Bellville, Texas; E-mail: jwebb@valmont.com
ABSTRACT ................................................................................................................................................ 3
INTRODUCTION ...................................................................................................................................... 3
DESIGN CONSIDERATIONS ................................................................................................................. 4
ACCESS .................................................................................................................................................. 5
FRAMING ............................................................................................................................................... 6
LOADS ..................................................................................................................................................... 6
FOUNDATIONS ...................................................................................................................................... 8
INSTALLATION ..................................................................................................................................... 11
CONCLUSION ........................................................................................................................................ 12
REFERENCES ......................................................................................................................................... 13
In 2016, LCRA TSC was in the process of upgrading their existing 69kV Goldthwaite to Lampasas
Transmission Line. A portion of the line between San Saba and Lampasas crosses the Colorado River.
Several structures were located in the predicted 100 year flood plain. In addition, the structures directly
adjacent to the banks of the river were potentially located in areas of higher water flow. The purpose of
this paper is to discuss how LCRA TSC, HDR, and Valmont all worked together to develop a hybrid H-
Frame structure, consisting of concrete bases and steel tops, which leverages the benefits of both concrete
and steel to provide an efficient structural design with limited conflicts to construction installation or
performance. The paper will provide insight on the design and construction considerations that took place.
INTRODUCTION
In 2016, the Lower Colorado River Authority Transmission Services Corporation (LCRA TSC) worked
with HDR, Inc. in 2016 to upgrade 33.4 miles of existing 69kV transmission line from Lampasas, TX to
San Saba, TX, see Figure 1. The line would begin at the existing Lampasas substation, enter and exit through
the existing Lometa substation and terminate at the existing San Saba substation. The project scope
included the complete tear down of the existing 69kV wooden H-frame transmission line, re-building it
with 138kV concrete and steel H-frames and single pole structures. This project utilized 795kcmil, “Drake”
ACSR conductor along with an Alcoa #52 OPGW and 3/8” HS shield wire.
Colorado River
Crossing
The project was designed and constructed within existing centerline easements. A blowout analysis of the
existing line was completed, and the width of the existing ROW width was determined. Many existing
easements had structure type limitations, which specifically described what type of structures, mostly H-
frames, could be utilized in the easement. In order to comply with these easement limitations, H-frame
dead-end structures were used. Concrete structures were utilized where terrain and easement access
allowed, reducing costs and providing improved structure resiliency against flood waters commonly
experienced in the region.
The structural design parameters utilized for the project included NESC (2012) Rule 250B for light and
heavy loading districts in addition to Rule 250C, 90mph extreme wind, and Rule 250D concurrent ice and
The upgraded 138kV transmission line crosses the Colorado River at a single location. The river crossing
has a 100-year flood plain that affects three (3) H-frame structures. This crossing required a miscellaneous
easement permit from the General Land Office.
The general site conditions along the project required that structures exceeding 100ft of overall length to
be made of steel due to access constraints and shipping effort. A simple analysis was complete and it was
determined that two concrete poles less than 100 ft in length could be shipped on the same truck. In addition
due to erosion, differences in grade elevation, access paths, and other restrictions, a site specific
investigation was performed along the project to ensure the feasibility of concrete pole delivery. In areas
deemed unfit for concrete pole delivery, steel poles were utilized.
Texas often experiences heavy storm events that cause an elevated risk for flooding, due to the soil
characteristics and topography. LCRA TSC has design parameters that help to limit the number of outages
caused by flood waters. The 100-year flood clearance requirement is used in bodies of water, or low lying
areas, where water may improperly drain, or has been shown historically, by rain gauges and rainfall
intensity duration functions, to produce excessive runoff.
This paper will discuss the design and construction considerations that took place on LCRA TSC’s T138,
138kV Transmission Line Upgrade Project, for hybrid structures 201, 202, and 203, used within the 100-
year flood plain of the Colorado River.
DESIGN CONSIDERATIONS
Optimum structure spotting was performed utilizing PLS-CADD during the preliminary design efforts of
the T138 transmission line rebuild. This optimum spotting effort led to the reduction in the number of
structures impacted from the 100-year flood plain elevation from four (4) to three (3).
However, back span structures 78/3, 78/4, and 78/5 were optimized based off clearance, span length, cost
for structure type, and restricted zones to eliminate the need for structure 78/4 while still maintaining
blowout clearances within the (ROW).
ACCESS
The utilization of concrete structures was primarily determined by the overall structure height and site
access for construction activities. Overall, the transmission line structures were comprised of several
variations of structures; including concrete and steel H-frames, steel single poles, and concrete 3-pole
deadends. Over 80% of the transmission line structures were concrete h-frames. This number of concrete
structures is indicative of the terrain of the right of way and the number of structures whose length was less
than 100 ft. However, the access was deemed unacceptable for the use of concrete structures near the
Colorado River crossing after a field visit by LCRA TSC and their construction managers. The reasoning
for this determination included such observations as elevation changes, soil competency, water
encroachment, vegetation and maneuverability of the shipping vehicle. In addition to access, this region
required taller structures exceeding 100 ft in length, to achieve the minimum clearances above the ground
and the 100-year flood elevation. The decision was made to utilize a hybrid structure comprised of concrete
pole base sections and steel pole tops, in order to facilitate structure delivery to the site, and to improve
durability, considering the base of the structure may be frequently inundated by flood waters. The concrete
pole base sections extended to 5 ft above the predicted 100 year flood elevation. It was assumed that
concrete sections less than 40 ft in length were able to be shipped to the site.
The hybrid concrete/steel pole H-frame structure, though different from the concrete or steel pole H-frames
used on the majority of the rest of the line, also meet the requirements of the existing easements that limited
the line to using H-frames. The hybrid pole framing can be found in Figure 4 below.
LOADS
Water flow assumptions would drive the effort in potential flow loads, impact loading, scour, and wave
loading experienced on the structures. LCRA TSC provided historical water data for a 100 year flood event
located along the Colorado River just upstream from the T138 transmission line crossing location at the
San Saba gauge. This data provided a flow rate of 224,000 cubic feet per second (cfs), which was used in
determining the flow velocity at the crossing location, as well as gauge datum and heights to derive the
flood elevation. Calculating the cross sectional area of the flow from the profile view of the PLS CADD
model and provided elevation of flood, it was determined that the average flow through the cross section
would be 2.78 ft/sec. However, this flow velocity needed further assumption across the flood region as it
is unreasonable that this average flow velocity remain consistent across the entire cross section. Due to the
variations in water depth and amount of vegetation present along the cross section, it was determined that
the water velocity at a distance away from the river bed would substantially decrease. HDR and LCRA
TSC assumed that the closest structure location, greater than 500ft from the everyday river bed location,
and due to the amount of vegetation in the area that there was negligible water flow at any of the hybrid
structure locations. This assumption allowed HDR and LCRA TSC to be comfortable with considering both
The structural design of the hybrid structure was governed by NESC (2012) with all applicable load factors
and design requirements. Supplemental code, API 2A-WSD, would be used to develop the flow loads
within the flood plain of the Colorado River.
NESC (2012) loads cases were used in the structural design of the hybrid structures. In addition, LCRA
TSC standard load cases and overload factors were applied. The loading conditions for which loads were
applied include the following: NESC Rule 250B Heavy, NESC Rule 250B Light, NESC Rule 250C, NESC
Rule 250D, Broken Wire, and everyday 60 degrees Fahrenheit.
Additional assumptions were also made for uplift. The hybrid structure framing would utilize direct
embedment foundations. The subsurface soil conditions present near the Colorado River were
predominately hard lean clay (CL) and very stiff sandy silty clay (CL-ML). Geotechnical investigations at
the structure locations were important to analyze the foundations in bearing and uplift due to the additional
hydrostatic water pressure caused by the flood. It was assumed that this soil profile would drain freely and
the entire hydrostatic force cause by the flood waters would be utilized in the design of these structures.
The design flood elevation provided by LCRA’s TSC historical 100 year water data was 1160 ft, which
translated to 20 ft of standing water at the design structure location. This pressure head would translate to
a substantial additional uplift force on the H-frame structure. Discussions with LCRA TSC, HDR, and past
project experiences with Valmont concrete poles, allowed for acceptable assumptions to be made on the
concrete pole base diameters. These assumptions included a ground line diameter of three feet (3 ft), a butt
diameter of four feet (4 ft) and a hollow shaft with thickness equal to nine inches (9 in). These assumptions
drastically reduced the amount of uplift pressure present at the base of the structure due to hydrostatic
pressure.
The flood load weather condition was provided by LCRA TSC. This extreme weather condition was said
to occur at 60 degrees Fahrenheit, 30mph wind, and no ice. Furthermore, wind driven waves would not be
considered from this weather condition due to the nature of the river flow, which is negligible. The
Exposure Category of the structure was taken to be L-3 API (2002), since the structures are unmanned and
have low consequences of failure due to the rural location and probability of extreme weather event.
Therefore, the loading criteria of the hybrid structures should consider hydrostatic forces, wind loading on
the structure, and water pressure forces. These loads were provided within the issued for price quote
drawings to Valmont in the form of a load tree.
Pressure forces were resolved into transverse point loads based off the Static Wave Analysis using the
Morison Equation, API (2002):
Where,
F = Hydrodynamic Force Vector per unit length acting normal to the axis of the member
F = Drag force vector per unit length acting to the axis of the member in the plane of the member axis,
U, lb/ft
Where,
d = Length of the submerged pole above ground = 20ft
= 1.0
Hydrostatic Pressures exerted on the pole will also be present as defined below:
=ϓ = 2215 /
Where,
F = Hydrostatic Pressure acting normal to the bearing area of the structure
ϓ = Unit weight of water, 62.4 lbs/ft3
L = 35.5ft (Depth of water ~20ft + embedment of 15.5ft of standard embedment)
And,
=( ) = (1.0) 9,460 ~9.5kip (Each Pole)
Where,
R = Hydrostatic Force acting upwards along the bearing area
π( )
A = = 4.27ft , bearing area structure
D = Effective outside diameter of the circular cylinder members ~4ft (assumed butt diameter)
"
d =D − = 3.25ft = Effective Inside diameter of the circular cylinder member
"
LF = 1.0
FOUNDATIONS
Foundation design of the hybrid structures followed LCRA TSC standard methodology utilizing Power
Line System module CAISSON (2016). LCRA TSC methodology includes independent leg design and
The directly embedded structures were designed to include a standard embedment equal to 10% of the total
pole length plus and additional 4ft. In addition, an increased depth of soil was neglected in the foundation
design to account for local scour. Select backfill was utilized for the hybrid structure foundations to limit
the possibility of local scour, and to reduce the bearing area exposed to the additional hydrostatic forces
that may otherwise be present if concrete backfill was specified. Special care was taken during the
foundation design to ensure the soil profile had enough embedment to limit the deflection and shear forces
exerted on the structure utilizing select backfill. Furthermore, fully saturated soil properties were utilized
for the required embedment determinations.
ADDITIONAL CONSIDERATIONS
In addition to the proximity of the structures to the Colorado River, there are also several airports in the
vicinity including, Lampasas Airport, San Saba County Municipal Airport, Lometa Air Strip and Birchfield
Ranch Airport. Therefore LCRA TSC was required to provide notice to the Federal Aviation
Administration of our proposed construction/alteration of the existing T138 San Saba to Lampasas
transmission line. It was determined that the proposed structure heights and locations did not pose a threat
to navigation, therefore and additional marking or lighting was not required.
MANUFACTURING PROCESS
Typically, concrete poles are delivered on the right-of-way to the structure staking location. Due to
extremely wet weather in the winter of 2016/2017 the contractor requested that the poles be delivered to a
marshalling yard to allow for delivery to proceed even though ROW conditions did not allow for just-in-
time structure delivery to the staking location. The steel top sections were also delivered to the marshalling
yard. Once conditions allowed for access to the river crossing location the contractor was able to move the
sections to the ROW using their equipment.
INSTALLATION
The installation of these hybrid pole H-frames was essentially the same as a standard H-frame, with a few
modifications.
One thing that was brought up early on in the process was maintaining the plumbness of the concrete
structures during erection. With the upper steel structure being completely assembled prior to erection,
some of the flexibility to adjust the structure was removed. It was essential to set the base of the structure
as best as possible, or the upper portion of the structure could possibly be off.
CONCLUSIONS
The use of hybrid structures adjacent to or in flood plains have been considered for several projects since
their use on this project. Once of the reasons the hybrid structures have not been utilized is because the
potential for flowing water has not been present. The design and drawings have been added to our standard
design library. We have also given presentations internally discussing this design.
For LCRA TSC there was not a substantial cost increase with the hybrid h-frame structure, from the type
of structure that would usually be used in such a flood zone. In the past LCRA TSC has installed anchor
bolted steel dead end structure with concrete pier foundations, that utilize an additional reveal of
approximately 10-15 ft. An anchor bolted structures would have both additional cost for the foundation
and the structure itself. In addition getting a concrete truck to the structure would also be difficult which
would also lead to increased cost. After all of these consideration LCRA TSC determined that the hybrid
h-frame would be the most effective.
LCRA TSC’s mission statement is “To enhance the quality of life of the Texas we serve through water
stewardship, energy and community service.” One way we do this is by increasing the reliability of our
transmission lines. We believe that utilizing the hybrid h-frame structure and providing a minimum of 25’
of clearance above the high water elevation are pertinent in providing reliability to our customers. The
hybrid h-frame structure utilizes concrete bases in areas where water could be present for extended periods
REFERENCES
NESC (2012) National Electric Safety Code, National Bureau of Standards, Washington, D.C.
API (2002) Recommended Practice for Planning, Designing and Constructing Offshore Platforms -
Working Stress Design, API Recommend Practice 2A-WSD, American Petroleum Institute, 21st Edition
200-2002.
The primary objective of this report is to present the technical background of AC Interference as
well as to provide a summary and conclusions specific to the Xcel Energy AC Interference
Program. This report addresses AC interference modes and their effects on metallic pipelines
with respect to corrosion and safety hazards. In addition, this report summarizes the
methodology behind the AC Interference program as well as identifies additional potential risk
factors.
Several authors contributed to the development of this report. They are as follows:
Mr. Ross is an employee of Xcel Energy Services Inc., the service company for the Xcel Energy
Inc. utility holding company system. XES provides engineering, siting and land rights, project
management and other services for the utility operating company subsidiaries of Xcel Energy
Inc., including Southwestern Public Service Company (SPS). In this paper, XES and SPS are
collectively referred to as “Xcel Energy”. Mr. Ross is employed in the Transmission business
unit of XES.
Contents
1 Introduction and Program Overview .................................................................................................... 1
2 Current Work Practices ....................................................................................................................... 2
2.1 Screening Criteria...................................................................................................................... 2
2.2 Notification ................................................................................................................................. 2
3 AC Interference Modes ....................................................................................................................... 2
3.1 Capacitive Coupling .................................................................................................................. 3
3.2 Inductive Coupling ..................................................................................................................... 3
3.3 Resistive Coupling..................................................................................................................... 4
4 AC Interference Program Methodology............................................................................................... 4
4.1 Task 1: Desktop Identification ................................................................................................... 6
4.2 Task 2: Field Delineation .......................................................................................................... 7
4.2.1 Pre-field Planning Effort ............................................................................................... 7
4.2.2 Field Delineation........................................................................................................... 8
4.2.3 Post-Field Data Processing and Development of Deliverables ................................... 8
4.3 Task 3: Initial Prioritization ........................................................................................................ 9
4.4 Task 4: Electrical Resistivity Data Collection ............................................................................ 9
4.5 Task 5: Final Data Gathering .................................................................................................. 10
4.6 Task 6: AC Interference Modeling........................................................................................... 11
4.7 Task 7: AC Interference Design Review ................................................................................. 12
4.8 Task 8: AC Interference Mitigation Construction .................................................................... 13
4.8.1 Surface and/or Near-Surface Grounding ................................................................... 13
4.8.2 Deep Grounding ......................................................................................................... 13
5 Additional Potential Risk Factors ...................................................................................................... 14
5.1 Increased or Decreased AC Interference Buffer Zones .......................................................... 14
6 Conclusions ....................................................................................................................................... 15
7 References ........................................................................................................................................ 16
The Southeast New Mexico (SENM) portion of the SPS System has a standalone transmission
expansion portfolio driven by major developments within the potash mining industry, increase in
oil and gas exploration and production efforts, and commercial growth within the area. SPS
worked closely with Southwest Power Pool (SPP) to develop cost effective solutions to address
the incremental oil and gas electrical load additions in the SENM region. In 2014, SPP approved
and issued Notices to Construct to SPS for the addition of 149 miles of 115kV electrical
transmission lines, and 245 miles of 345kV electrical transmission lines.
Throughout the years working within the SPS System, we have dealt with many challenges
specific to the existing oil and gas infrastructure within this area. Traditionally, our challenges
have stemmed from land access, easements, and load growth specifically related to the oil and
gas operators within the area. However, we now have a new challenge to consider, that being
issues surrounding potential AC Interference with coated metallic structures and specifically
buried metallic oil, gas, and water pipelines.
Recent studies conducted by oil and gas industry groups (i.e. Interstate Natural Gas Association
of America and American Gas Association) as well as corrosion engineering industry groups
(i.e. National Association of Corrosion Engineers) have provided documentation and examples
of the effects of AC Interference when electrical transmission and/or distribution infrastructure is
located adjacent to, crosses, and/or runs parallel to metallic structures. The AC Interference is
caused by the electromagnetic, capacitive, inductive, and conductive effects from the electrical
transmission and/or distribution infrastructure. In certain situations, this interference can cause
electrical arcing, step-current interaction and/or accelerated corrosion damage to metallic
structures such as above and below ground pipelines, metering and/or regulator facilities, well
heads, and storage tanks.
Specific to our Power for the Plains build-out within the SPS System, we have started a program
to better understand the potential effects of AC Interference caused by new electrical
transmission line projects on metallic oil, gas, and water pipeline infrastructure as well as to
define Xcel Energy’s risk profile specific to this topic.
When a new transmission line project is kicked off, the Engineers are aware to take
underground pipelines in to account when starting the design. If the project is in an area with
potential risk for AC Interference, then they reference the screening criteria. Within the criteria is
a buffer value that is used as an initial definition of extents during the process of investigating
the presence of metallic pipelines for a given electrical transmission line. Using different
evaluation criteria from industry published documents helps determine which pipelines identified
could potentially experience AC Interference.
2.2 Notification
If the screening procedure listed above identifies metallic pipelines that could have a potential
risk for AC Interference, the transmission engineer coordinates with the Sighting and Land
Rights agent to send a notification letter to that pipeline owner. The intent of the letter is to make
the owner aware of the construction that will take place and provide the owner with a point of
contact within Xcel Energy if they have questions.
3 AC Interference Modes
If the right set of circumstances exist, electrical interference from electrical transmission lines
can occur and may affect metallic structures in a variety of ways including developing shock
potential at or around a given metallic object and/or causing accelerated corrosion of the
metallic object. There are three generally accepted primary modes of AC interference (or
coupling):
Capacitive Coupling;
Resistive Coupling.
The paragraphs below provide a brief description of each type of AC interference mode.
How capacitive coupling works is a metallic structure (like a metallic pipeline) builds a charge as
a capacitor with the surrounding air acting as the dielectric, which can maintain the electric field
with a minimum loss in power, resulting in a potential difference with the surrounding earth. The
magnitude of potential is primarily dependent upon the pipeline proximity to the electric
transmission line, the magnitude of electric transmission current, and the individual phase
arrangement. If the potential buildup due to capacitive coupling is significant, electrostatic
interference may present a risk of electric shock and/or arcing.
The inductive effects of an electrical transmission line on an adjacent buried metallic structure
are a function of geometry, soil resistivity, coating resistance of the buried metallic structure,
and the operating parameters of the electrical transmission line. The geometry characteristics
include separation distance between the pipeline and the electrical transmission line towers,
depth of cover for the buried metallic structure, angle between the buried metallic structure and
the electrical transmission line, and phase conductor configuration. These parameters remain
relatively constant over the service life of both the electrical transmission line and the buried
metallic structure. The coating resistance, power system resistance, and soil resistivity could
vary with the seasonal changes and as the structures begin to age. The operating parameters
of an electrical transmission line, such as phase conductor load, phase balance, voltage, and
available fault current, which all contribute to inductive coupling, can vary significantly. An
individual conductor current load and phase balance are dynamic and changes with load
requirements and switching surges. These variations in operating parameters contribute to
variations in levels of inductive coupling. During normal electrical transmission line operation,
the current load varies as the load demand changes both daily and seasonally. While normal
operating conditions are often referred to as steady state, the term is somewhat misleading as
the current loads and corresponding inducted AC potentials can be continuously varying, adding
further complexity to quantifying inductive coupling.
During fault conditions, the primary concern is the resistive coupling being transferred through a
soil mass to a buried metallic structure. However, it should be noted inductive coupling can also
be a concern as the phase current and corresponding inductive coupling will most likely
increase. What this means is during a fault condition, the inductive effects during normal
operation increase due to the elevated electromagnetic interference during the fault period.
In addition to the tasks themselves, we needed to define the initial extents of our study for each
SPS System project. To develop these extents, we and HDR completed a literature review
specific to AC Induction which included studying reports commissioned and/or published by the
following:
A detailed list of the reports we examined during our literature review is provided in the
References section of this report. In addition to reviewing reports authored by others, we also
discussed this topic with HDR’s Corrosion Engineering Team as well as two other Corrosion
Engineering Firms, those being the Wood Group and Southern Cathodic Protection.
What we found consistent amongst all of the information we reviewed was the distance of
impact due to AC interference which was possible between an electrical transmission line and a
metallic pipeline varied markedly from project to project but relied on three basic elements:
Soil conditions (i.e. whether a given soil mass was conductive for electrical
interference and if so, to what extent?); and
As such, to set initial extents or “buffers” for each SPS System project, we utilized those
recommended by INGAA. It should be noted while we recognize these initial buffer values to
likely be conservative, they are buffer values which are generally accepted by the oil and gas
pipeline industry as well as the corrosion engineering industry for conducting initial AC Induction
modeling. For reference, we are considering using the following buffer values:
Aerial Photography
In addition, HDR has access to the IHS Energy’s and PLATTS Pipeline Databases, which are
proprietary, subscription-based databases that provide additional pipeline and well pad data. As
such, for each project, HDR reviews data within the IHS Pipeline Database and includes any
corresponding pipeline and/or well pad data into the desktop maps. Also, any data which was
captured by Xcel Energy’s Siting and Land Rights group and/or project surveyors is included in
the initial desktop maps. At the end of Task 1, HDR produces desktop maps which include the
following:
transmission line);
The produced maps are considered a 50% Draft map package and are provided to Xcel Energy
in Adobe (.pdf) ArcGIS (.shp), AutoCAD (.dxf), and Google Earth (.kmz) formats.
It has been HDR’s experience that while the utility locate One Call ticket identifies and flags the
majority of underground pipelines and utilities along the proposed electrical transmission route,
occasionally in remote or high-density areas, a pipeline is not identified by the utility locate One
Call ticket. Typically, these unidentified pipelines are older and at times, abandoned assets,
having been installed several decades earlier. These pipelines are often the most important to
locate, as pipeline engineering standards of the time did not require cathodic protection, and the
steel used in construction is usually of lower quality than what modern-day standards require.
Within the transmission line ROW, other aboveground and belowground utilities such as high-
density polyethylene (poly) pipelines, telephone cables, fiber optic cables, are also
recorded. Outside of the transmission line ROW, the locations of poly lines and other utilities
are not recorded since they do not pose the same conductivity issue as metallic pipelines.
In areas where the presence of a pipeline is suspected, typically due to the identification of a
linear scar on aerial photographs, but no identifying information such as utility placards or pin
flags are available, a commercially available handheld utility locator is utilized to determine
whether a buried metallic structure is present. The Radio-detection Model 7100 utility locator is
used in several different modes to allow for reasonable certainty as to whether a buried metallic
object is present within a particular area.
It should be noted sub-meter accuracy is only present at the locations where the GPS points
were collected. HDR’s field crews make every effort to collect as many points as possible
where a pipeline may turn or bend, but surface indicators of a point of inflection are not always
present, and therefore may not be accurately captured. This generally is not a concern within
the transmission line ROW, as the majority of the pipelines are marked with pin flags, but
outside of the transmission line ROW this can become an issue. HDR can only be confident in
the accuracy of the location of a pipeline where a GPS point was collected; the real-world
locations of pipelines and the inferred lines digitized in GIS to represent them may not correlate
exactly in a given location.
Once the pipelines have been digitized, the maps generated in Task 1 are updated to only
reflect metallic pipelines and other utilities which were captured during the field delineation. The
maps are then reviewed for quality control, and become Attachment A of a Technical
Memorandum. The Technical Memorandum identifies all metallic pipelines, well pads, or other
above and below-ground utilities along the route and lists them by closest electrical
transmission structure name or by mile post if electrical transmission structure locations are
unavailable. Occasionally, the presence of a pipeline is suspected due to some piece of data,
but unable to be identified in the field. These areas are highlighted on the maps and in the
Technical Memorandum as areas “Recommended for Further Investigation”.
HDR’s deliverables to Xcel Energy include the Technical Memorandum, and digital maps, which
can be considered a 95% map package, in GIS (.shp), AutoCAD (.dxf), and Google Earth
(.kmz). The Technical Memorandum includes the Pipeline Inventory Maps and a list of pipeline
and well pad operators and contact phone numbers collected along the transmission line
route. The final step is to upload all deliverables to the corresponding Xcel Energy electrical
transmission line ProjectWise folder.
Separation Distance: distance between the electrical transmission line and a given
pipeline;
Colocation / Crossing Angle: angle at which a given pipeline crosses and/or parallel’s an
electrical transmission line;
Colocation Length: distance a pipeline is within the corresponding buffer distance for a
given electrical transmission line;
Pipeline Owner: interstate oil and gas transmission operators versus midstream
operators
Planning access routes and location groupings to most efficiently perform the testing;
Mobilizing personnel and equipment to the site, performing quality equipment checks
including resistance test box calibration check;
Collecting Wenner Array electrical resistivity data in general accordance with the
American Society of Test Materials (ASTM) Standard Test Method G57 and IEEE
Standard 81-2012 along two perpendicular arrays. Each array will consists of “A”-
spacings which include 0.5, 1.0, 3.0, 5.0, 10.0, 15.0, 20.0, 30.0, 50.0 and 100.0 feet.
Performing onsite data quality checks including; standard deviation/error reporting for
internal meter readings, deviation check between arrays and data plotting for log/log
curve smoothness; and
Pipeline diameter;
In-line inspection and testing records (which provide an indication as to the remaining
Similarly, Xcel Energy is sent an RFI to obtain physical and operational parameters of the
proposed electrical transmission line. During the modeling process, if additional information is
needed, supplemental RFI(s) are developed and sent to the corresponding project stakeholder.
As expected, not all pipeline owners are able to provide all of the requested information. When
this occurs, assumptions about the unknown information are made. Prior to executing Task 6:
AC Induction Modeling, all input parameters are confirmed and agreed to by all project
stakeholders (which include all pipeline owners and Xcel Energy). The software used for this
modeling has the ability to print the input data of all three basic elements (electrical transmission
line, soil resistivity, and pipeline characteristics) so both the known data provided and assumed
input values are shared to confirm their validity.
The CDEGS ROW system model for a given project is constructed by using typical cross-
sectional configurations of the corridor in question. This includes all transmission and metallic
pipeline facilities. The ROW model can analyze the system under steady state, fault, and
lightning strike conditions. Additionally, the soil type can vary throughout an electrical
transmission line alignment, which is something the HiFREQ model does not easily allow.
The CDEGS HiFREQ system model is built using the same information as the ROW model.
Unlike ROW, HIFREQ allows for above ground conductors (valve locations in this case) to be
modeled. Additionally, HiFREQ allows more modeling options for energizing conductors. The
limitation of HiFREQ is the limitation of a single soil structure for the entire electrical
transmission line alignment, resulting in the possible need for multiple models to be developed.
After completing the CDEGS modeling, HDR compiles the results in a report with sections for
each specific pipeline associated with a given Xcel Energy electrical transmission line. These
results are the basis identifying areas of AC Induction risk along a given metallic pipeline asset
as well as developing strategies for mitigating the AC Induction risk.
Using the CDEGS software, modeling the interaction of an electrical transmission line’s steady
state, peak, and fault condition effects on nearby metallic structures in a given media (soil
resistivity), as described in Task 6, provides the AC voltage and current predictions along the
length of a metallic pipeline. If AC voltage is shown to exceed 15V at a given location, USDOT
Pipeline and Hazardous Material Safety Administration (PHMSA) and National Association of
Corrosion Engineers (NACE) design guidelines dictate mitigation methods must be utilized to
dissipate AC effectively while preventing valuable cathodic protection current (DC) loss, all the
while maintaining safety. The specific PHMSA and NACE design guidelines are as follows:
Examples of mitigation methods may include but are not limited to utilizing discharge locations,
isolation, continuity, decoupling, equalization, and dissipation, all of which are elements which
need to be designed specific to a given pipeline asset.
For this program, potential AC Mitigation design is being led by the individual pipeline
companies and/or their consultants because any mitigation would typically be installed on the
pipelines rather than the electrical transmission line. Through this process, Xcel Energy and
HDR will be reviewing the AC Mitigation designs developed by others to confirm their
effectiveness in reducing voltage and current to acceptable levels as prescribed by PHMSA and
NACE.
Deep Grounding
For project sites which have surficial soils that display a high resistivity, one downside for
surface and/or near surface grounding is the length of mitigation ribbon and/or the number of
shallow anodes which may be required to achieve the desired mitigation results. This method
situation can prove to be problematic in areas where rights-of-way (ROW are constrained such
as in urban areas or ROW’s which contain multiple utilities.
In general, construction costs are generally higher for deep well grounding than for comparable
surface and/or near-surface grounding. However, deep well grounding can be a viable option in
specific applications where one or both of the following criteria are true:
The soil resistivity at the ground surface is significantly greater than the soil resistivity at
lower depths; and/or
Horizontal surface grounding is not feasible due to construction constraints (i.e. limited
ROW)
For typical AC induction mitigation systems, where surface and/or near surface and deep
grounding are both viable options, surface and/or near surface grounding generally proves to be
more efficient and economical because it can achieve a lower resistance to ground for lower
overall cost. The table below provides a comparison of three different mitigation options: zinc
ribbon and copper cable, which are both surface to near surface mitigation options, and the
deep well, which is a deep grounding mitigation option.
Effective in Effective
Effective Soils with
Ease Cost History in Areas
Mitigation as Extremely
of of of with
Option Fault High
Construction Construction Effectiveness Resistivity Limited
Mitigation
Values ROW
Zinc
Moderate Low Yes Likely Unlikely Unlikely
Ribbon
Copper
Yes Moderate Yes Likely Likely Unlikely
Cable
Deep
No High Yes Unlikely Likely Unlikely
Well
These values were utilized as our buffer values because they are generally accepted within the
Oil and Gas and Corrosion Engineering industries. However, we now know the impact of AC
interference from an electrical transmission line onto a metallic pipeline is influenced by soil
resistivity, voltage of the electrical transmission line, geometry between the electrical
transmission line and a given pipeline, and the existing cathodic protection system on a given
pipeline. As such, there could be scenarios where the utilized buffer zone underestimates the
actual impact distance of AC Interference. For example, in the situation of a 345 kV electrical
transmission line, which is constructed in extremely high conductive soils conditions and which
parallels a metallic pipeline for several miles, the AC Interference impact distance could extend
greater than 1,500 feet.
6 Conclusions
Given the continued transmission build out within the SPS region, this program was tasked to
investigate the potential impact to new transmission line projects sited near metallic pipelines.
After more than a year of investigation, there is one clear finding: AC Interference from high
voltage transmission lines onto metallic pipelines needs to be analyzed in order to be
understood. With the continued construction of electrical transmission lines within the SPS
region, as well as the dense existing infrastructure of the Oil & Gas industry within this same
area, it is not practical for Xcel Energy to site its transmission lines to completely avoid all risk
specific to AC Interference.
Historically, AC Interference has not been a “hot topic” within either the electrical or oil and gas
industries. However, given the increased events (i.e. ruptures, leaks, etc.) associated with the
oil and gas pipeline industry over the last several years, there is now Federal regulation reform
within the oil and gas pipeline industry which highlights AC Interference and associated AC
Corrosion.
Energy electrical transmission line projects on existing pipelines to ensure we are taking
reasonable measures to understand and address AC interference.
7 References
B. Tribollet, “AC Induced Corrosion of Underground Pipelines”, Underground Pipeline Corrosion
Conference, June 2014.
CAN/CSA C22.3 No.1 – 06 “Overhead Systems”, Standards Council of Canada, July 2007.
EL-904 Research Project 742-1 “ Mutual Design Considerations for Overhead AC Transmission
Lines and Gas Transmission Pipelines, Volume 1: Engineering Analysis”, Electric Power
Research Institute (EPRI), September 1978.
EL-3106 “Power Line Induced AC Potential on Natural Gas Pipelines for Complex Rights-of-Way
Configurations, Volume 1: Engineering Analysis”, Electric Power Research Institute (EPRI) and
American Gas Association (AGA), May 1983.
Giddings, J., Marshall, D., and Moore, C., “Shocking News – Analyzing AC Inducted Potentials on
Buried Pipelines”, American Society of Civil Engineers (ASCE) Pipelines Conference, Portland,
Oregon, May 2014.
Gummow, R.A. “A/C Interference Guideline Final Report”, Canadian Energy Pipeline Association,
June 2014
INGAA Report No. 2015-04, “Criteria for Pipelines Co-Existing with Electric Power Lines”,
Interstate Natural Gas Association of America (INGAA), October 2015.
Johar, U.M. and Shwehdi, M.H., “Transmission Line EMF Interference with Buried Pipelines:
Essential and Cautions”, Proceedings from the International Conference on Non-Ionizing
Radiation, October 2003.
NACE TG327, “AC Corrosion State-of-the-Art: Corrosion Rate, Mechanism and Mitigation
Requirements, National Association of Corrosion Engineers (NACE) Report 35110, October
2010.
Keywords—Temporary Protective Grounding, Temporary III. SOURCES OF ELECTRICAL HAZARDS WHEN WORK IS
Protective Grounding Equipment, Equipotential Zone, PERFORMED ON DE-ENERGIZED LINES AND EQUIPMENT
Transmission Operations
A. Accidental Closing of a Circuit’s Isolating Device
I. NOTICE Procedures to de-energize and isolate lines and equipment
The content of this paper is for discussion purposes only are well established in the electric utility business and are
and is not intended to be a substitute for the work and safety required by the U.S. OSHA [6] when work is to be performed
policies and procedures that may be in place or appropriate for on de-energized lines and equipment. These procedures
other companies. The views expressed therein are those of the include a means of ensuring that the lines and equipment are
authors and not intended to be advice or substitutes for other not energized until the work is completed and the workers are
companies’ rules, etc. or advice as to compliance with U.S. in the clear of the lines or equipment. Failure to follow these
procedures can result in the intentional but erroneous closing of
OSHA rules and regulations.
the circuit’s isolating device.
F. Lightning
The hazards at a worksite associated with lightning are V. BODY CURRENT EXPOSURE LIMITS
significant but difficult to predict. The use of TPG procedures
and equipment may not be sufficient to protect workers from A. For Cases of Accidental Energization with Exposure Time
those hazards [8]. of 0.03 s to 3.0 s
Equation (1) indicates that the body current limit for
preventing ventricular fibrillation is a function of the time to
which a worker is exposed to current flowing through their
body. In the case of accidental energization of a de-energized
C. Isolation from the Hazardous Differences of Electrical XI. ENGINEERING ANALYSIS METHODOLOGY FOR
Potential DETERMINING THE LEVEL OF HAZARDOUS DIFFERENCES IN
Some companies use the term isolation different than it will ELECTRICAL POTENTIAL
be used in this paper. For purposes of this paper, isolation is An engineering analysis is necessary to create TPG
the practice of isolating a worker or the public from sources of procedures and methods that will protect workers from
hazardous differences in electrical potential that can occur hazardous differences of electrical potential. This analysis
when de-energized and grounded work is being performed. must include determining the fault current and its X/R ratio at
The restricted area must be large enough to exclude the area of the worksite and the clearing time for sources of accidental
hazardous levels of ground potential rise. Examples include: energization. This information can be provided through the
use of fault modeling software in use at most electric utilities.
• Restricting access to grounded structures where TPG
This information is typically readily available for substation
equipment has been installed.
buses due to the need for such values for calculating the
• Restricting access to locations where TPG equipment settings of protective relay systems, sizing fuses, and
has been installed at some distance away from a calculating arc flash exposure levels. Therefore, the analysis is
structure and a temporary ground rod is used. fairly straight forward for worksites at substations. However,
for worksites on transmission lines the information may not be
• Restricting access to grounded vehicles or work readily available and may require significant resources to
equipment. obtain this information for the worksite for any planned or
unplanned work. Factors to be considered for the engineering
X. PURPOSE OF TEMPORARY PROTECTIVE GROUNDING analysis methodology are discussed below.
A. Limit Worker Exposure Voltage A. Use of Impedance Versus Resistance for TPG Equipment
Workers engaged in de-energized and grounded work must The exposure voltage across a worker when working on a
be protected from hazardous differences of electrical potential grounded line is a function of the voltage drop (4) across the
that can occur at a worksite due to accidental energization or TPG equipment that is installed in parallel with the worker
induction. This protection is provided by TPG equipment [26]. The exposure voltage limit from (3) can be used in (5) to
installed on lines and equipment to limit currents through a determine the maximum impedance of the TPG equipment
worker by limiting the exposure voltage at the worksite [24]. such that the exposure voltage limit will not be exceeded. The
The proper application of TPG equipment will create a low reactance component of the impedance is a function of the
impedance path in parallel with the worker that will limit the geometry of the installation of TPG equipment. On some
body current to acceptable levels. worksites the configuration changes as the work progresses. It
can be a difficult task to determine the impedance unless
B. Limit the Exposure Time modeling software is used for each configuration. A simplified
Equation (1) indicates that levels of body current that may approach to take into account the inductive reactance is given
result in ventricular fibrillation are a function of the time that a in [27] where correction factors have be developed to
worker is exposed to the current. For cases of accidental approximate the inductance as a function of the cable
energization of de-energized lines and equipment the proper resistance of a limited number of TPG equipment
use of TPG equipment will ensure that the fault clearing configurations. This approach works fairly well for substation
protective device associated with the source of accidental worksites where the TPG equipment configuration is more
energization will clear the fault as quickly as possible. For predictable for the different phases of the work. It is more
cases of accidental energization of two or three phases this is difficult to apply this method for line work where the
accomplished by using TPG equipment to create a multi-phase configuration of the TPG equipment is dynamic as the work
short circuit that is connected to the best ground electrode at progresses at a worksite. Some utilities use only the resistance
the worksite. For cases of single-phase accidental of the TPG equipment for both line and substation work and
energization, there must be a low impedance fault current include a safety factor to compensate for not using the
return path that will ensure sufficient current flows during the impedance.
fault to operate the fault clearing protective devices as quickly
By:
Abstract ......................................................................................................................................................... 3
1. Background ............................................................................................................................................... 4
2. Clearance Considerations.......................................................................................................................... 4
4. Special Considerations and Limitations for Ultra-Tall Tubular Steel Poles ............................................. 8
7. Conclusion ............................................................................................................................................. 14
8. References .............................................................................................................................................. 15
Abstract
Due to Houston’s rapidly growing population, the Texas Department of Transportation (TxDOT) is
consistently rebuilding and adding roadways to their extensive network of highways. Transmission lines
crossing these expanding TxDOT roads are often required to be raised or relocated to accommodate the
expanding infrastructure. At some TxDOT interchanges with highly dense vertical stacking of roadways,
transmission structures can exceed 200 feet in height in order to meet both National Electric Safety Code
(NESC) and TxDOT clearance requirements.
Tubular steel poles (TSP) have been selected for several recent projects involving TxDOT expansions
requiring new crossing structures of 200 feet or more. TSPs have a smaller footprint when compared to
the traditional lattice tower structures. As these projects are often located in parts of the city with very
limited space to spot the new structures, a single foundation for a TSP is more favorable than the four
foundations and wider base required for a lattice tower structure. Additionally, TSPs result in shorter
construction times as they require less assembly when compared to lattice towers.
Although TSPs are quickly becoming the preferred structure type for these applications, they do not come
without their own set of considerations and design limitation challenges. This paper will explore key
design concepts and lessons learned from installing TSPs 200 feet and taller, such as setting deflection
limits, cambering vs raking, constructability concerns, foundation design considerations, and Federal
Aviation Administration (FAA) lighting requirements.
1. Background
Texas is home to 10 of the top 15 fastest growing cities in the United States with populations of 50,000
residences or more according to the United States Census Bureau[1]. Number one on the list is Conroe,
Texas, a northern suburb of Houston, Texas. Newly released United States census shows that Conroe had
a growth rate of 11 times that of the United States from 2015 to 2016. Other list toppers are Frisco and
Georgetown, Texas, suburbs to Dallas and Austin respectively. With the spike in population and growth
seen in these Texas cities and their suburbs, huge demand increases for resources and infrastructure are
required to sustain the rapidly growing population.
Like any major metropolitan area in the United States, a large percentage of employment is located in the
heart of Texas’ major cities, such as Houston and Dallas. While corporation’s offices are located
downtown, most of the employees reside in the suburbs and commute to work daily. Due to overly
congested roadways from this rapidly growing population, the Texas Department of Transportation
(TxDOT) is required to increase the capacity on their roadways by adding to or expanding the existing
interstate and highway systems in these rapidly growing areas.
To help alleviate congested highways, rehabilitate bridges, and improve connectivity in and around the
state’s metropolitan areas, Texas voters approved $5 billion in bonding. The Texas legislature authorized
the bonding in 2007, under Proposition 12[2]. The first $2 billion in projects were approved in 2010 and
the remaining $3 billion was approved in 2011. In 2015, voters approved Proposition 7[3], which allocated
portions of the state’s revenue to the State Highway Fund for non-tolled projects. These substantial
budget allocations have allowed for large scale and fast-paced projects which are often centered on
expanding existing interchanges, but, due to the established city build-up in these areas, there is often
limited physical space for any new transportation infrastructure. As a result, new roadways connecting
suburbs to city centers typically need to be stacked vertically at interchanges.
Existing transmission lines, which often run parallel to major roadways, frequently cross these congested
interchanges. As TxDOT projects result in the vertical expansion of interchanges, existing transmission
line crossings at or near these interchanges are also required to be raised. To meet National Electric Safety
Code[4] (NESC) clearance standards and TxDOT clearance requirements, the new designs can result in
structures that exceed 200 feet in height. Structures exceeding 200 feet in height are referred to as ultra-
tall structures throughout this paper.
This paper is focused on how the rapid increase of population affects transmission line design
considerations in highly congested major Texas cities. This paper explores the methodology for designing
transmission lines in highly congested areas with limited physical space that require ultra-tall structures.
2. Clearance Considerations
In designing transmission lines crossing over interchanges in Texas’ major metropolitan areas, many
challenges arise. The lack of available physical space and the high costs associated with rerouting
transmission line alignments often make raising them the best option to deal with the increasing
interchange heights.
For transmission lines, determining the correct electrical clearance requirements for a given line is the
first major design consideration to tackle. The NESC is often the basis for determining the minimum
clearance requirements for transmission lines in Texas. In some cases, state or local codes can be more
stringent, therefore it is important to investigate all project specific requirements prior to finalizing a
clearance criteria. For the purposes of this paper, the NESC and the Texas Administrative Code[5] (TAC)
will be assumed as the only applicable clearance requirements.
NESC Rule 232 provides the vertical clearances for communication cables, shieldwires and conductors to
roads, streets and other areas subject to truck traffic. While NESC Rule 234 provides the clearances for
communication cables, shieldwires and conductors to bridges. The NESC clearances for Rules 232 and
234 for phase to phase operating voltages of 0 kV, 34.5 kV, 69 kV, 138 kV and 345 kV are summarized
in Table 1.
Table 1. NESC Rules 232 & 234: Wires, Conductors, or Cables to Bridges, Roads, Streets and
Other Areas Subject to Truck Traffic
Communication
Shieldwires (ft)
34.5 kV (ft)
138 kV (ft)
345 kV (ft)
69 kV (ft)
Nature of surface underneath wires, conductors, or
(ft)
cables
Roads, streets and other areas subject to truck traffic 18.5 18.5 18.5 19.1 20.6 24.8
Bridges - where wires are attached 5.5 5.5 5.5 6.1 7.6 11.8
Bridges - where wires are not attached 12.5 12.5 12.5 13.1 14.6 18.8
Bridge structure - where wires are not attached,
7.5 7.5 7.5 8.1 9.6 13.8
accessible to pedestrians
Bridge structure - where wires are not attached, no
6.5 6.5 6.5 7.1 8.6 12.8
pedestrian access
In addition to the general electrical gap clearances over bridges and roads, the NESC also specifies a
requirement to account for the effects of the electric field for voltages above 98 kV phase to ground.
NESC section 232D3C, states that the electro static effects on the largest expected vehicle to cross the
electric filed are to be limited to 5 mA or less when the transmission line is operating at 120 degrees
Fahrenheit. Depending on the voltage, conductor type and configuration, this rule has the potential to
control the design clearance for a given interchange crossing.
The TAC Section 21.41 Overhead Electrical and Communication Lines provides the TxDOT clearance
requirements of overhead electrical and communication lines, under maximum operating temperatures, to
roadways and bridges owned and operated by the State of Texas. Table 2 summarizes the requirements
outlined in Section 21.41.
Table 2. TAC Section 21.41 Overhead Electrical and Communication Lines
Communication
Shieldwires (ft)
34.5 kV (ft)
138 kV (ft)
345 kV (ft)
69 kV (ft)
cables
After the minimum clearances to the roadway are accounted for, an additional height buffer is typically
added to accommodate potential changes in the TxDOT design that could result in increasing the final
height of the new roadway as well as field changes during the installation of the transmission line and
construction tolerances. This is important because the transmission work is typically completed well in
advance of the final roadway designs and the buffer helps to mitigate potential redesign efforts including
re-fabricating structure portions which can be very costly depending how far along the structure is in
production and the extent of the changes that are required. This buffer may also be increased to account
for the construction equipment needed to build the interchange. The TxDOT construction crews may not
be certified to work around electric circuits and so larger approach distances (see OSHA 29 CFR, §1910.
269[6]) may be needed between the construction equipment and the electrified conductors.
In addition to the roadway height, code clearances and buffers, the configuration, conductor and line
voltage will also add to the final structure height. Adequate phase spacing is required and will vary based
on the voltage class of the transmission line. Conductor sag can increase or decrease depending on the
conductor size and type used as well as the maximum ampacity rating of the line. The structure
configuration is often based on a mix of the special limitations in the area as well as the utility
preferences.
Roadway on and off ramps can be stacked two to three decks high to allow uninterrupted traffic at
interchanges. Typically, clearance between road decks is 20 feet or more with deck thickness at
approximately eight to ten feet. Transmission structures are spotted as close to the roadway as possible to
avoid crossings around the maximum sag of the conductors, but may still encounter ten feet of sag or
more. Assuming 345 kV vertical configuration with phase spacing of 25 feet with shield wires above the
top phases, structure lengths can quickly exceed 200 feet or more. Figure 1 illustrates the potential
required pole height at an interchange with 3 road decks stacked vertically.
Figure 1. Example Clearance Factors*
* The dimensions shown here are general and do not necessarily represent
minimum clearances or spacing. It is the responsibility of the project designer
to verify and determine these dimensions based on each specific project’s
parameters.
Clients often request transmission lines to be raised using the existing conductor types and operational
voltages but require the structures to be designed to handle an increased numbers of circuits, different
attachment configurations and/or an increase in voltage class. Future buildout considerations can
significantly increases the height of the structures.
the multi-section design approach, TSPs are easy to ship on standard trailers and straightforward to
construct. Some of the constraints placed on the pole design can result in very expensive structures, but
the overall design benefits often outweigh the potential cost adders. For these reasons, TSPs are not only
the more favorable option, but sometimes they are the only option.
mills in America have the capacity to mill some of the larger size base plates and often they have
extensive back logs that can extend lead times well beyond 26 weeks. In these instances, the base plate
size may be reduced by using multiple bolt circles, however, this practice can create additional issues
since the inner bolt circle is often located on the inside of the pole. To tighten the anchor nuts inside the
pole, a worker must have access to the interior of the pole which often means a special access door must
be designed and fabricated into the side of the base section. Special detailing for these requirements adds
cost and production time to the structure order. It is best to discuss any potential procurement and design
concerns with perspective TSP vendors as early in the design process as possible to avoid critical delays
later in the project schedule.
Not all considerations are critical to the structural design or construability of a project. For projects that
are located along heavily trafficked areas, pole aesthetics can play a large role in how the public reacts to
a new transmission installation. Pole deflection is often the most visually apparent of these concerns
especially on ultra-tall TSPs. For instance, Figure 2 shows that even a minor deflection of five percent
results in over 12 feet of deflection at the tip of a 250 foot TSP. Although this amount of deflection may
not be a structural issue, it is visually identifiable and may raise concerns from the public. To control the
deflection and alleviate potential concerns, a designer can specify a deflection limit, design stiffeners,
design a pole camber or rake the pole during installation.
Figure 2. Visual Tubular Steel Pole Deflection
As shown in the figure above, deflection in TSPs is nonlinear. The tip deflection is best approximated
using the deflection equation for a cantilevered beam:
=
3
Where:
δ = deflection
P = the resultant load at the structure tip
L = the length of the pole
E = the modulus of elasticity of steel
I= the moment of inertia of the pole itself
For a given project, P, L and E are all constant, therefor adjusting I becomes the only way to vary the tip
deflection. For projects with a set deflection limit, the equation can be rearranged to determine the
required moment of inertia:
=
3
In practice, the desired value of I can be achieved by adjusting the section diameters and/or increasing the
thickness of the section walls. The option to limit deflection will generally cause an increase in costs or
result in larger base diameters, potentially making a pole design unsuitable for a project. Designing pole
stiffeners along the length of the pole shaft to increase the moment of inertia of the pole may be another
option to control deflection, but this option will also increase costs.
Cambering is also used to counteract the visual effects of deflection without making special adjustments
to the overall pole design. Cambering is the process of fabricating the TSP with a preset curve to offset
anticipated deflection that will occur once the load is applied. If applied correctly, the pole will appear
plumb under typical “every day” conditions after all the wires are installed and tensioned. However,
fabricating curved sections can increase structure cost and it is difficult to accurately camber the pole to
match the actual installed loading. Because cambering is not an exact science, after the pole is installed
and placed under load, the applied load may not be high enough to “straighten” out the camber, the
distribution of the camber along the pole may not be even, or the pole camber may be less than
anticipated and the pole ends up deflecting toward the ultimate loading regardless. The results of an
unsuccessful camber design is a higher cost pole that still shows a visual deflection.
Raking is the process where the pole is tilted away from the direction of anticipated deflection before the
pole is placed under load. The pole can be raked by adjusting leveling nuts below the baseplate on the
anchor bolts. The amount of rake is usually based on the anticipated deflection under normal everyday
loading, but its success heavily depends on the installation crew both installing the conductors to the
specified tensions and tilting the pole to the correct angle. If the pole is not cambered, a straight, raked
pole will still deflect and appear curved along its length. Poles of this length can still be visually
unappealing due to the deflection.
Beyond the aesthetic effects, deflections this large can also create issues with the wire sag and tension.
During construction, the pole will deflect as each wire is brought up to the specified design tension. If
each wire is tensioned individually, each successive wire being strung will result in an increase in the
overall pole deflection. This in turn will cause the reduction in tension (increase in sag) for each of the
already tensioned wires. This effect is illustrated below in Figure 3. To avoid potential clearance issues
after the wires are installed, it is important to discuss this effect with the construction crews and verify
that the final sag/tension measurements were checked after all the wires were installed.
requirements if the structures are located proximal to an airport. The specific requirements for each
project are detailed in the permits once they are approved by the FAA and are generally required to be
installed as well as active during the construction process.
The constructability of a design is a critical aspect that must be considered during the initial project
planning. Items such as site access, equipment weight limitations, current versus future grading plans and
clearances to the surrounding environment both above and below ground have to be factored into the
design. When exploring potential structure locations, the available physical space above ground is not the
only concern. Foundations must be placed so as to avoid underground conflicts, such as pipelines or other
burred infrastructure. To identify these potential obstructions, underground investigations should be done
prior to starting any engineering as finding them during construction can result in costly delays and
extensive redesign efforts. To avoid constructability issues in the field, it is best to discuss design
decisions with the construction team early on, as they are able to provide insight on how they will stage
and execute the build process.
Transmission line construction often requires a temporary or staged work plan to help accommodate
limited line outages, shorten overall construction schedules or to accommodate a phased construction of
the roadway. Although temporary construction is a typical practice in transmission line work, it is
especially crucial for projects involving the permanent installation of ultra-tall TSPs as construction over
major roadways must not create significant disruption to the daily traffic patterns. For higher voltage or
critical lines, outages cannot be scheduled for a long duration and usually can only be arranged for
specific times in short durations not long enough to remove existing structures and completely install the
permanent structure and foundation. Temporary structures are generally set slightly off of the existing
alignment and the existing conductors transferred from the existing structures to the temporary locations
during scheduled short outages. Once the conductors are properly installed in a temporary configuration
and the outage expires, the transmission service is restored. This allows the construction of the permanent
ultra-tall TSPs on the existing alignment to continue while the line remains in service. Although
temporary designs may not need to handle the final loading, they still must physically fit within the given
area of the rebuild while not obstructing the accessibility of construction equipment during the installation
of the permanent structures.
Depending on the specific project scope, installing all new conductors and wires may not be ideal and
instead reusing the existing cables could be preferred. When transferring conductors and wires as part of a
rebuild, additional lengths must be spliced in to maintain the specified catenaries and tensions. Once the
splices are made, typical wire pulling methods are utilized to string the line. It is crucial for the
construction crew to understand the intended tensions for both the final configuration and any temporary
work so that wire length and splice locations can be properly arranged to avoid rework between the
construction stages. The designer must also provide enough spare cable and splices in the bill of material
to accommodate the construction process as well as the final intended design. Not accounting for the
additional wire or splices can create shortages in the field resulting in construction delays and re-work or
potentially even cause deviations from intended design and excessive line tension.
Installing ultra-tall TSPs often will cause uplift issues on the adjacent structures, this is due to the abrupt
increase of height between the existing and new structures. Depending on the span lengths and the
difference in height, several adjacent structures may need to be replaced in order to create a gradual rise
in the line that eliminates uplift on adjacent tangent assembles. To properly estimate a construction
schedule and budget, it is important to review the structures beyond the immediate rebuild area for uplift
concerns as replacing adjacent structures will increase the cost and construction duration of the overall
project significantly.
Long term maintenance issues should also be considered during the design process. To perform
maintenance items such as inspections, hardware and conductor work, repainting rusted portions of the
pole with galvanizing paint and replacing FAA lighting will require very extremely tall bucket lifts or
other specialized equipment. In situations where this equipment may not be readily available to the utility
crews, it may be possible to add special design details to the pole such as permanent ladders and tie off
points that can be used for future maintenance operations.
7. Conclusion
As the population in Texas continues to grow, increasing the demand on TxDOT roadways, the need for
ultra-tall transmission structures will continue to become more and more common. Similar growth
patterns are occurring in many other metropolitan areas through the country. For these extremely
congested areas with limited rerouting options, rebuilding transmission lines around TxDOT interchanges
will require structures with footprints that minimize ground and aerial conflicts. TSPs are being utilized
more frequently for these applications as they are better suited to meet the design constraints when
compared to other traditional structure options like lattice towers. Although TSPs may provide the best
solution for a given project, they still have many limitations and require a great deal of consideration by
the designer. To achieve critical milestones and eliminate costly delays, project stakeholders such as
TxDOT personnel, design team members, construction leads, and the TSP vendor should meet at the early
stages of the project and continue regular meetings throughout. This open and continuous communication
between the project members will ultimately lead to a more effective and successful installation.
8. References
[1] “The South Is Home to 10 of the 15 Fastest-Growing Large Cities.” US Census Bureau, 25 May
2017, https://www.census.gov/newsroom/press-releases/2017/cb17-81-population-estimates-
subcounty.html
[2] “Proposition 12 (General Obligation Bond Projects).” Texas Department of Transportation,
http://www.txdot.gov/inside-txdot/projects/prop12.htm. Accessed 22 June 2017.
[3] “Proposition 7 - Constitutional Amendment for Transportation Funding.” Texas Department of
Transportation, http://www.txdot.gov/government/legislative/state-affairs/ballot-proposition-
7.html. Accessed 22 June 2017.
[4] National Electric Safety Code C2-2017, Institute of Electrical and Electronics Engineers. 26 April
2016.
[5] “Texas Administrative Code Title 43.1.21.C Rule 21.41.” Texas Registrar,
http://texreg.sos.state.tx.us/public/readtac$ext.TacPage?sl=R&app=9&p_dir=&p_rloc=&p_tloc=
&p_ploc=&pg=1&p_tac=&ti=43&pt=1&ch=21&rl=41. Accessed 22 June 2017.
[6] “Occupational Safety and Health Standards 1910.269.” US Department of Labor,
https://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=STANDARDS&p_id=986
8. Accessed 22 June 2017.
This paper focuses on the basics of substation grounding as well as providing back ground on step over
native soil (outside the station). With this knowledge the decision of whether to study and mitigate for
these conditions can be made by individual utilities.
INTRODUCTION
IEEE Std 80-2013 focuses on the safety of personnel within a substation. Calculations are performed to
evaluate a substation grounding system for IEEE Std 80 compliance against two measures, touch voltage
and step voltage.
Touch voltage is the voltage difference between a person’s feet and an object they touch. Step voltage is
the voltage difference between a person’s feet.
When evaluating a grounding system for IEEE Std 80 compliance, a tolerable touch and step limit is
determined. This tolerance limit takes factors such as the backup clearing time, soil properties, and any
material between the grounding system and the person’s feet. That compliance limit is then compared
to the calculated touch and step value. A grounding system can be declared IEEE compliant if the touch
Grounding plays an important role throughout the entire power system, protecting equipment and
personnel from normal and abnormal system conditions. IEEE Std 80 is the IEEE Guide for Safety in AC
Substations Grounding and is used to provide guidance on limits based on scenarios presented in a
substation environment and behavior of the human body when subjected to an electric shock. Through
an understanding of the body’s response to electric shock, calculations can be performed to determine
compliance, or safety, limits for both step and touch voltage.
The resistance of the grounding system can be referred to as the resistance to remote earth as shown in
the figure below.
Transmission Line
SLG Source
Fault Substation
(I)
Resistance
to Remote
Earth
(R)
The graph below is a representation of the potential on the surface of the earth as the distance from the
station increases. While the exact rate of decay and magnitude of the potential will be different for each
station depending on the fault magnitude, grounding system resistance, and soil type, it is known that
the largest difference in potential will be seen right outside of the substation.
Analysis Methodology
The analysis of a grounding system’s performance under fault conditions requires these basic steps:
• Characterizing the soil and developing a model that represents an electrical equivalent of the
soil in which the grounding system is installed.
• Determining the amount of current that will flow through the substation grounding system to
return to the source through the earth.
• Determining IEEE Std 80-2013 touch and step voltage compliance levels.
• Analyzing the actual touch and step voltage values throughout and three feet beyond the
substation.
The following sections discuss the general methodology of each step. Multiple software packages are
used throughout the industry. This paper does not focus on a specific software or analysis tool.
IEEE Std 81-2012 (“IEEE Guide for Measuring Earth Resistivity, Ground Resistance, and Earth Surface
Potentials of a Grounding System”) indicates that “probe spacings on the order of the ground grid
dimension should give sufficient model accuracy for computing touch and step voltages”. If the soil
resistivity data maximum probe spacing for a substation is much less than the maximum substation
dimensions, the grounding analysis accuracy may be impacted.
Faulted Substation
The maximum fault duration, a.k.a. backup clearing time, can be calculated or a conservative value
assumed. Commonly, 0.5 seconds (30 cycles) is used as typical maximum fault duration for transmission
voltages. As the maximum fault duration decreases, the compliance limits increase.
Soil surfacing can have the greatest impact on the compliance limits. Surfacing varies from crushed
gravel to asphalt and many other options. Each surfacing will have its own electrical properties
depending on the material and the thickness. While the intent of the substation surfacing is to insulate
personnel from the ground, the type and depth of material will vary based on utility needs and the
function of the substation. For example, driving a vehicle in 6 inches of crushed gravel is often not
feasible, so a maximum of 4 inches is used or a different surfacing is applied for driveways. Occasionally,
there may be areas of substations where there is no surfacing; this is referred to as step-over native. The
overall compliance limits are a function of both the resistivity of the surfacing and the thickness. As the
thickness increases, the compliance limits will increase; as the resistivity of the surfacing increases, the
compliance limits will also increase.
Since step voltage is the potential difference between a person’s feet, any current would flow through
the person’s feet and leg. As noted in IEEE Std 80 Clause 16.1, “step voltages are inherently less
dangerous than [touch] voltages” based on the fact that the majority of the body current will not flow
through the heart (Clause 7.2); yet the guide does not provide details on how the compliance limit could
be increased to account for this aspect. It should also be noted that touch potential violations are far
more common than step.
Certain conditions may increase the likelihood of having step over native issues. One such condition is
where the soil has high resistivity and few return paths for the fault current distribution. Another is
when the soil model reveals a low resistivity layer on top of a high resistivity layer. Below is the example
from a station in western Texas where step over native issues were calculated. As can be seen, the
second layer of soil has a significantly higher resistivity than the first.
Due to the high resistivity of the second layer, the voltage will remain near the surface longer since the
path of least resistance is in the top layer of soil. The plot below shows the potential along the surface
from the station fence going out 1,000 feet. With a station fault current of 10,000 Amps applied, the
GPR at the station is as 15,000 volts. At 500 feet from the station, there is still a GPR of over 3,000 volts.
Not only is that a high GPR, but the rapid decay of potential on the surface leads to step voltages in
excess of the IEEE compliance limits.
Due to the vague nature of the standard at this time, it is up to the individual utility to determine what
their best practice is.
While step over native concerns are uncommon, they can be costly to mitigate. One of the quickest
options for mitigating step over native is to extend the surfacing father than normal. Due to property
lines, and topography that is not always an option. It is not uncommon for stations with issues to see
that the station surfacing would need to be extended as far 15 and 20 feet from the station fence.
Extending the grounding system can be the most effective mitigation method. An outer grounding loop
is often installed 3 feet beyond the substation fence. Extending the grounding system would be adding
additional grounding runs deeper than the rest of the grounding system. This method is referred to as
stair stepping. The intention is to push the voltage further into the earth, away from the surface. This
option is also depend on the property lines and often not an option.
An additional mitigation technique is to take measures to decrease the station grounding resistance.
That can be accomplished by adding additional grounding, ground rods, and ground wells. Enhancing
fault current distribution
When step over native issues are discovered, any combination of the above mentioned mitigation
techniques can be used. While mitigation can occasionally be simple, it can also be extravagant. In the
example from above, analysis techniques were employed to reduce the fault current by applying fault
current split and using a realistic fault clearing time as opposed to a conservative value. Even with a
reduced fault current and faster clearing time, step over native issues were still present. The image
below is the output from grounding software where the color areas denote non-compliance.
To mitigate the step over native issues, an additional ground loop was added around the outside of the
station and a total of seven (7) 80 ft deep, 6 inch diameter ground wells were added to the station.
The grounding analysis indicated step over native concerns directly outside of the fence extending as far
as 25 feet from the station fence. Due to property limitations, extending the surfacing and adding
additional loops outside of the station were not an option. The image on the left shows the existing
station grounding while the image on the right is the ground grid design required for step over native
compliance.
The following was required to bring the station to IEEE Std 80 compliance including step over native
outside of the station fence:
• Approximately 9,000 feet of existing 500 kcmil bare stranded copper conductor buried 18 inches
below structural grade.
• Approximately 24,000 feet of new 500 kcmil bare stranded copper conductor buried 36 inches
below structural grade.
• Approximately 7,200 feet of existing 4/0 bare stranded copper conductor buried 18 inches
below structural grade.
• Approximately 280 feet of #4 AWG bare stranded copper conductor for the connection of the
fence to the grid
• 331 ground rods (194 new), each 5/8 inches in diameter and 10 feet long.
• Shift northern fence in 3 feet and taper corners for step voltage over native compliance
considerations
1. Substation Transformers
Figure 1: A typical substation showing circuit
breakers and high voltage bus work. Substation transformers are important asset in an
electric substation. These are essentially static
Besides asset condition, CenterPoint Energy also devices that operate at primary voltages ranging from
replaces substation assets proactively based on 12 kV to 345 kV. They are used to step down
capacity constraints. System planning engineers at electricity from high voltage down to a lower voltage
CenterPoint Energy conduct steady-state, short level that is more appropriate for distribution to local
circuit and dynamic analyses based on load areas. They consist of insulated windings installed
forecasting. Based on results of such planning studies on laminated steel cores and immersed in oil that
the company continually and routinely replaces insulates and cools the equipment.
substation equipment to maintain safety and grid
reliability. As three-phase banks, transformer assemblies
can include either three single-phase units or one
CenterPoint Energy has developed in-house asset three-phase unit. Transformer assemblies consist of
life cycle systems. These systems provide following key components:
intelligence in the form of analytics to asset
management, operations and engineering to help Core and Coils
prioritize assets for replacement based on condition Insulation & Bushings
and system risk. Since timing of this work is usually Tanks
not driven by a specific event, such as a projected Cooling system
summer – peak demand, this work is sometimes
De-energized/On-Load Tap changers
shifted forward or backward by few months to meet
outage windows and manage resources needed to Foundations & spill containment
address time critical work. However, occasionally it Protection & Controls
becomes necessary to replace failed substation
Transformers have no preset life expectancy, but
equipment that has resulted in a significant outage or
utilities generally expect transformers to have a
has created an unsafe condition. Such replacements
useful life span of 40 to 50 years, assuming normal
are performed on an emergency basis just prior to or
loads and operating conditions. In a typical electric
immediately after an asset failure.
system, transformers operate under many extreme
conditions, both normal and abnormal, which affect
Substation equipment, such as a power
their aging and breakdown. Overloads cause above
transformer or a circuit breaker, typically has an
expected life span of 40 years or even longer if normal temperatures, through-faults can cause
maintained properly. In contrast, the life span of displacement of coils and insulation.
2. Mobile Transformers
Council on Large Electric Systems (CIGRE) Figure 9: Typical mix of circuit breakers based on
studies on circuit breaker reliability have confirmed operating medium, at all transmission voltages
that mechanical degradation presents greater end-of-
life concerns than electrical degradation. Even for In order to prioritize breakers for replacement,
modern SF6 gas breakers, mechanical degradation CenterPoint Energy conducts evaluation of its assets
causes 80% of observed problems. Generally, based on maintenance assessment that includes
operating mechanisms, bearings, linkages and drive following factors:
rods represent components that experience most
mechanical degradation problems. Gas, oil and air Bus fault levels & misoperations
leakage also occur. Newer classes of breakers use Tank and breaker type
simpler operating mechanisms and have seals and O&M expense & Maintenance Activity
fittings that help minimize leaks. Shown below is a No. of Operations & vintage
typical relationship between number of years a Operating mechanism & manufacturer
beaker remains in-service and the unit maintenance
cost.
b) Solid-State Relays
Some of the parameters that could be considered Predictive maintenance based on analytics
during development of analytics based spare strategy involves using software technology for real time
include: monitoring of equipment condition and comparing its
current operational state to its initial state under
spare levels and their strategic locations normal and emergency operating conditions e.g.
asset installed base & its condition predictive maintenance strategy can be developed for
system impact and lead times a substation transformer based on infra-red
compliance requirements, if any inspections. Infra-red model can be developed and
O&M costs and failure rates then compared using advanced algorithms with the
initial state model to determine any variances or
Future failure rates can be calculated based on invisible thermal signatures of impending damage
past historical failure information and condition before the damage occurs such as rapid deterioration
based asset health scores [6]. of insulation.
VIII. REFERENCES
IX. BIOGRAPHY
By:
PAGE
Abstract ......................................................................................................................................................... 3
1. Introduction .............................................................................................................................................. 3
6. References .............................................................................................................................................. 15
While the programs’ main difference is the modeling of the drilled pier behavior (MFAD as rigid and
LPile as flexible member), this study focused on how each program accounted for the interaction of the
foundation shafts with the surrounding subsurface media. As there is a wide variety of subsurface
materials and conditions with various characteristic distinctions, a review of existing p-y (lateral load
transfer) models and non-linear spring parameters was completed to determine the appropriate soil
parameters to be utilized for the analysis. Multiple foundation shaft sizes, with varying aspect ratios, were
analyzed for a combination of different subsurface material engineering properties and structural loads
using both LPile and MFAD. The performance of each of the foundation designs was assessed based on
the strength and deflection criteria.
By comparing the resulting designs, this study provides basic guidelines to help determine which program
would yield the most optimized solution for a given set of parameters and eliminate potentially over-
conservative designs.
1. Introduction
Drilled shaft and direct embedded foundation types have been widely utilized to support transmission line
structures subjected to line loads. In general, any transmission line structure can be categorized into a
rigid or rotational structure based on the system’s resistance to deformations. In contrast to rigid
structures such as lattice towers and braced H-frames where uplift, compression and shear demands are of
significance, foundations of rotational structures such as monopoles and unbraced H-frames need to be
resisted for large overturning moment and shear demands. While direct embedded foundation designs are
mostly limited to lightly loaded tangent pole structures, drilled shaft foundations are more applicable for
resisting heavier transmission line loads. This study is limited to the analysis and design of drilled shaft
foundations of rotational structures (e.g.: steel monopoles).
There is a variety of software tools that perform the analysis and design of drilled shaft foundations of
rotational structures (moment-governed), of which, MFAD and LPile are the most commonly utilized
packages in the transmission line industry. The design outputs of both foundation software packages are
dependent on a set of user-specified inputs that can be categorized based on geotechnical conditions,
geometry of the shaft and applied loads. Based on the structural specifications of different utilities and the
geography of the project-site, these input parameters can fluctuate significantly, thus resulting in vastly
different output designs. The primary objective of this study is to assess the drilled shaft foundation
design outputs generated by MFAD and LPile for a given set of controllably varied input parameters.
2. Design Methodologies
It is crucial to have an understanding of how the subsurface material and shaft capacities are computed
within each program in order to determine the appropriate applicability and degree of conservatism. As in
the case of foundation engineering, soil-structure interaction becomes significant because of the
difference in mobilization of the structure with respect to the soil. Both LPile and MFAD utilize nonlinear
Winkler-spring mechanisms to model the appropriate nonlinear response of the subsurface material.
These responses are then calibrated using results of full-scale lateral load tests.
MFAD uses a four-spring load-deflection model to reflect the interaction between the drilled shaft and the
surrounding subsurface, as shown in Figure-1 below.
Figure 1- Free-body Diagram (left) and Schematic Diagram (right) [5] of Spring Model used in MFAD
These four types of springs relate the resisting forces acting on the drilled shaft with the resulting
deformations, as follows:
Recently, reliability-based design methods have been developed for different designs in foundation
engineering to account for uncertainties that exist in loading, geometry and geotechnical aspects of any
foundation. The traditional deterministic approach to account for such uncertainties is to include
conservative values of uncertain quantities and safety factors which often yield over-conservative designs.
Figure 2- Combined Probability Density Function representing limit state of a system [3]
In reliability-based design, g is the limit state function that divides the design space into two regions, the
safety region (g > 0) and the failure region (g < 0). The limit state function corresponds to the safety
margin measured as a difference between resistance, R, and load effect, Q, as illustrated in Figure-2.
Because of the input uncertainties, g in itself, is treated as a random variable with an assumed normal
distribution.
B. LPile
i. Spring Models
The subsurface surrounding the pile is modeled as a set of nonlinear springs in LPile to reflect that the
soil resistance p is a nonlinear function of pile deflection y.
Figure 3- Free-body Diagram (left) and Schematic Diagram (right) of Spring Model used in LPile
Unlike MFAD, the subsurface model in LPile only utilizes lateral translational springs for different
subsurface material layers, as shown in Figure 3. The p-y curve corresponding to the top translational
spring is to indicate that the pile is subject to a finite deflection with no subsurface resistance.
Derivation of p-y curve models in LPile is based on the results of full-scale foundation load tests. During
a load test on a laterally loaded pile, strain gauges are installed along the length of the pile that records
strains at gauge locations. From the direct measurement of strain and curvature, bending moment values
can be further derived. Evaluation of mobilized subsurface resistance along the length of the pile requires
two differentiations of bending moment curves with respect to depth. The experimental p-y curves are
then plotted showing the distribution of deflection and subsurface resistance.
For the implementation of the p-y method, the beam-column differential equation is solved in LPile using
the finite differences analysis method. The foundation shaft is divided into n intervals with a node at
Following the full-scale load testing procedure described above, strain values are recorded for increments
of the applied structural loads. The results of the full-scale load tests determine the boundary conditions
for each of the nodes of the n intervals. Applying these boundary conditions, the program evaluates a
condition of static equilibrium and computes the shear, moment and lateral deflection at each interval.
Problem Statement
A double circuit 150 feet, 90 degree, dead-end 138 kV steel monopole is considered in a National
Electrical Safety Code (NESC) light loading district. Maximum base reactions provided by the vendor
shall be utilized for the drilled shaft design. Consistent shaft geometry is used in this design example as
tabulated in Table 1. Geotechnical input parameters that are subject to variation as parts of this example
are tabulated in Table 2. Monolithic soil profiles with depths of 75 feet for cohesive and cohesion-less soil
types are utilized. Soft clay (Matlock) p-y curve is selected In LPile for poor clay profile. Stiff clay
without free water (Reese) is selected in LPile for average and good clay profiles. Sand (Reese) is
selected in LPile for all sand profiles. A constant water table depth of 33 feet is assumed for all profiles.
Maximum
Maximum Maximum Maximum Total Total
Input Pier
Pier Internal Moment Shear Rotation Deflection
Variation Internal
Moment (k-ft) Depth (ft) Depth (ft) (deg) (in)
Shear (k)
28757 13.1 940.6 42.8 2.13 17.8
Poor Clay 28324 8.9 707.5 39.6 0.68 5.6
28582 9.8 791.5 39.4 0.84 6.9
28563 12.4 916.5 42.8 0.44 2.8
Average Clay 27903 6.2 656.6 38.8 0.23 1.9
28073 6.5 732.4 37.1 0.29 2.2
27833 6.9 1061.2 31.1 0.19 0.6
Good Clay 28120 7.7 699.0 40.2 0.06 0.5
28131 7.3 762.0 38.1 0.07 0.6
LPILE
MFAD
MFAD - Springs OFF
Table 3- Results for Variation in Cohesive Soils
0 0 0
-20000 0 20000 40000 -1500 -1000 -500 0 500 -20 0 20
-10 -10 -10
Depth (ft)
Depth (ft)
Figure 4- Moment (left), Shear (center) and Deflection (right) Plots for Poor Clay Profile
Depth (ft)
Depth (ft)
Depth (ft)
Figure 5- Moment (left), Shear (center) and Deflection (right) Plots for Average Clay Profile
0 0 0
-20000 0 20000 40000 -1500 -1000 -500 0 500 -1 -0.5 0 0.5 1
Depth (ft)
Depth (ft)
Figure 6- Moment (left), Shear (center) and Deflection (right) Plots for Good Clay Profile
Maximum
Maximum Maximum Maximum Total Total
Input Pier
Pier Internal Moment Shear Rotation Deflection
Variation Internal
Moment (k-ft) Depth (ft) Depth (ft) (deg) (in)
Shear (k)
28931 12.4 794.2 39.3 0.27 1.2
Poor Sand
28741 10.5 757.4 43.7 0.24 2.2
28896 10.5 828.2 42.3 0.29 2.6
28730 11.0 845.7 33.1 0.23 0.9
Average Sand
28632 10.3 745.6 42.5 0.12 1.1
28757 10.2 808.0 41.4 0.15 1.3
28464 9.0 1004.3 28.3 0.19 0.6
Good Sand
28543 10.0 718.6 40.7 0.06 0.5
28627 9.9 775.0 39.5 0.07 0.5
LPILE
MFAD
MFAD - Springs OFF
Table 4- Results for Variation in Cohesion-less Soils
0 0 0
-20000 0 20000 40000 -1000 -500 0 500 -2 0 2 4
Depth (ft)
Depth (ft)
Figure 7- Moment (left), Shear (center) and Deflection (right) Plots for Poor Sand Profile
Depth (ft)
Depth (ft)
-40 -40 -40
Figure 8- Moment (left), Shear (center) and Deflection (right) Plots for Average Sand Profile
0 0 0
-20000 0 20000 40000 -1500 -1000 -500 0 500 -0.5 0 0.5 1
Depth (ft)
Depth (ft)
Figure 9- Moment (left), Shear (center) and Deflection (right) Plots for Good Sand Profile
Problem Statement
A single circuit, 150 feet tangent 345 kV steel monopole is considered in an NESC light loading district.
Maximum base reactions provided by the vendor are utilized for drilled shaft design. A consistent average
clay soil profile is considered for this example as documented in Table 5. A constant pier length of 30 feet
inclusive of a two feet reveal is considered for this design example.
Geotechnical Inputs
Depth Unit Weight Internal
Undrained
to Angle of
Layer p-y curve selection Shear
Bottom Total Effective Friction
No. in LPile Strength
of Layer (pcf) (pcf) (degrees),
(ksf), c
(ft) ϕ
Soft Clay
1 5 90
(Matlock)
Soft Clay
2 6 120 0.8
(Matlock)
Soft Clay
3 18 125 1
(Matlock)
Stiff Clay w/o
4 53 Free Water 125 62 2.5
(Reese)
5 69 Sand (Reese) 120 57 32
Stiff Clay w/o
6 75 Free Water 125 62 1.5
(Reese)
Loading Inputs
Load Shear Axial
Moment
Case Load Case Name Load Load
No. [kips] [kip-ft] [kips]
1 Vendor Outputs 16.436 1358.17 14.727
Table 5 – Geotechnical and Loading Inputs for Geometric Variation Design Example
Diameter of Shaft
Shaft Aspect Ratio Pier Length (ft)
(ft)
AR = 10 3 30
AR = 5 6 30
AR = 2 15 30
Table 6 - Geometric Variation
LPILE
MFAD
MFAD - Springs OFF
Table 7- Results for Variation in Aspect Ratio
0 0 0
-1000 0 1000 2000 -200 -100 0 100 -2 0 2 4
-5 -5 -5
Depth (ft)
Depth (ft)
Figure 10- Moment (left), Shear (center) and Deflection (right) Plots for Shafts with Aspect Ratio of 10
-5 -5 -5
Depth (ft)
Depth (ft)
Depth (ft)
-20 -20 -20
Figure 11- Moment (left), Shear (center) and Deflection (right) Plots for Shafts with Aspect Ratio of 5
0 0 0
-1000 0 1000 2000 -200 -100 0 100 -0.5 0 0.5 1
-5 -5 -5
Depth (ft)
Depth (ft)
Figure 12- Moment (left), Shear (center) and Deflection (right) Plots for Shafts with Aspect Ratio of 2
2. Performance-based observations depict vast differences in the total deflection and rotation values
calculated by LPile and MFAD. LPile typically calculates larger performance outputs than
MFAD. This is due to LPile’s primary modeling assumption of considering a flexible behavior of
a drilled shaft which, in contrast, is assumed as rigid behavior in MFAD.
3. The effect of springs is demonstrated by observing the close resemblance of results in clay
between LPile and MFAD with springs turned off. When springs are turned off in MFAD, side
shear, base shear and base moment contributions are eliminated and only lateral pressure
contribution is in effect. This is very similar to the LPile program’s methodology of solely
considering lateral pressure contribution in calculating soil capacity.
4. Comparing results in clay and sand profiles depict large conservatism in clay type p-y curve
models considered in LPile. We can infer that the p-y curve for ‘Sand’ model selection (Reese) in
LPile comply with the empirical test results utilized in MFAD. Cohesive soils develop more side
friction resistance than cohesion-less soils which has an apparent effect on the conservatism.
5. A very high aspect ratio of 10 (upper limit in MFAD) yields lower deflection results in MFAD
compared to LPile which treats the shaft as rigid. In contrast, a very low aspect ratio of 2 (lower
limit in MFAD) yields lower deflection results in LPile due to the limiting length of the shaft
which restricts the flex displacement of the physical pile. An average aspect ratio (4-8), which is
typically observed for transmission line structures, yields comparable results in MFAD and LPile
pertaining to its respective design philosophy.
6. References
[1] Nikos Gerolymos, George Gazetas. Development of Winkler model for static and dynamic
response of caisson foundations with soil and interface nonlinearities. Greece; 2005.
[2] David Bonhoff, Ph.D., P.E. Modeling Soil Behavior with Simple Springs. USA.
[3] Samuel G. Pakowsky. Standards and Reliability Based Design. USA.
[4] William M. Isenhower, Ph.D., P.E., Shin-Tower Wang, Ph.D., P.E., L. Gonzalo Vasquez, Ph.D.,
P.E. Technical Manual for LPile 2016. USA; 2016.
[5] Electric Power Research Institute, FAD Tools, FAD 5.1 User’s Guide. USA; 2015.
[6] Donald P. Coduto, Foundation Design Principles and Practices Second Edition. USA; 2001.
INTRODUCTION
Commissioning of substation security systems commonly
WHAT IS COMMISSIONING?
starts just weeks before the system is to be handed over
to the security operation center. But that can be a mistake. The commissioning process is designed to verify
that a substation security system operates in a
An integrator that doesn’t pre-test a system prior to manner consistent with the design documents.
commissioning, for example, may find the number of It involves testing all the security points on a
punch list items to be overwhelming, often preventing system and validating the integration between
the system from passing the commissioning test. security systems. This reduces the risk of system
errors and minimizes or avoids time and capital
Experience suggests that outcomes are far better expenses associated with fixing problems.
when commissioning is treated as an ongoing process
that spans multiple stages in a security system’s
design and installation. That process begins during
the security system design phase with the development should also address how the item integrates with other
of a commissioning plan. Following this step-by-step security systems.
approach is a recipe for successful outcomes with minimal
disruptions during the process. Site and floor plans — These plans should identify
the initial placement of and field of view for each security
camera. Those details will help an agent determine
STEP ONE: THE COMMISSIONING PLAN
if the cameras are installed correctly.
The security designer or commissioning agent can start
creating commissioning forms and schedules as soon as
Network schedule — This document, which lists the IP
the substation security system design is issued for bid or
address for each security device and piece of equipment
construction, depending on the project delivery method.
attached to the network, should be distributed to the
owner’s IT department to coordinate IP addresses,
These documents provide a commissioning agent with
subnets and other details. This schedule also can be
the tools needed to precheck the security system during
used to help verify if existing switches have the needed
factory testing and other early testing opportunities.
capacity or if additional network switches are needed.
to it. Some devices, such as fence detection systems the commissioning checklist. Proper cable management
and vehicle gate motors, can’t be installed in a factory and device installation should be verified during each visit.
test setting. The integrator should create scenarios
that simulate alarms for these devices. Once installation is complete, the commissioning agent
should verify that the as-built drawings and operation
The integrator also should set up a security workstation and maintenance manuals correspond with the
to verify that each point on the security equipment commissioning checklist.
matches the device shown on the workstation.
STEP FOUR: SECURITY
After individual device testing is complete, integration SYSTEM PRE-TESTING
testing between the security systems begins. This involves Pre-testing is performed to make sure a security system
testing alarm points to verify that the correct cameras are is operating properly so that acceptance testing can
activating, alarms are properly labeled and other security begin. This testing is typically performed by the integrator,
systems are initiating the correct response to an alarm. with the commissioning agent making periodic site visits
to the substation to observe.
STEP THREE: SECURITY
SYSTEM INSTALLATION Pre-testing begins with troubleshooting the system to
A smart commissioning agent will likewise conduct confirm that all security devices have been connected
periodic site visits during security system installation. to the correct points on the security equipment, and
The agent can use these visits to verify that security that all programming is finalized. A sizable portion of
equipment, devices and cabling are being installed troubleshooting issues are traced to incorrect labeling
according to the construction documents. While of cables and minor programming errors. A mislabeled
the agent’s role during these visits is to advance motion detector or camera, for example, will bring up
the commissioning process, rather than to perform the wrong location when an alarm activates.
construction administration duties, it is the agent’s duty
to inform the design team if it appears that guidelines Other tasks performed by the integrator during pre-
are not being followed. testing include cleaning equipment cabinets, equipment
filters and security devices so they perform optimally.
During installation, the agent also should verify that the By this stage, the integrator should use the final
equipment, cabinets, devices and cabling are labeled as commissioning checklist to perform a commissioning
prescribed in the design documents, which should match test of the entire security system.
During this phase, the security system alarm report is used Installation verification — The commissioning agent
to create a punch list. The integrator uses this list to verify should have verified that equipment, devices and
false alarms using video the system has recorded. If an cabling was properly installed during installation.
alarm is determined to be a nuisance, the integrator can During acceptance testing, the commissioning agent
take corrective action to solve the issue. should perform spot checks of the installation.
Punch list items should be resolved as they occur. If issues Penetration tests — Time permitting, the commissioning
persist, the owner should consider extending the burn-in team should perform random penetration tests during
period until the system is ready for acceptance testing. both day and night to find vulnerabilities an attacker
could exploit.
STEP SIX: ACCEPTANCE TESTING
Acceptance testing — the final tests before a security If the integrator and commissioning agent have worked
system is turned over to the owner — should be scheduled well together in resolving issues throughout the project,
by the integrator and commissioning agent prior to the acceptance testing should be relatively simple and
burn-in phase. At a minimum, the commissioning agent, trouble-free.
owner representative and integrator should be involved
in the final acceptance test. Most work will be performed CONCLUSION
from the substation, where they will activate alarms, verify Commissioning a substation security system is no small
device installation and perform other tests. One person is undertaking. The theoretical simplicity of checking things
typically stationed at the Security Operations Center to off a list can belie the importance of the process. Getting
verify that the system is functioning properly. better results and performing the process efficiently is
best achieved by making commissioning an ongoing
While performing the acceptance test, the commissioning element of the project.
agent should verify alarm naming conventions, map and
floor plan symbol locations, and network connectivity. Taking a step-by-step approach, beginning with a
Acceptance testing should include: comprehensive plan during design development, lays
a foundation for successful commissioning. Multistage
Alarm testing — Each alarm should be tested by testing, installation and burn-in time set the table for
simulating the conditions that would set off the alarm. acceptance testing, owner satisfaction and timely
To verify the alarm, the agent should confirm that it calls handover of the system to the owner.
up the correct camera. In some situations, more than one
camera may be used to verify alarms. BIOGRAPHY
Security system device verification — The agent should TERRY HARLESS is a senior physical security
verify that card readers, cameras, fence detection and specialist at Burns & McDonnell, with more than
other security devices are operating properly. 18 years’ experience in electronic security system
design and commissioning. In addition to writing
Camera views and fence zones — The agent should specifications for commissioning processes and
compare camera views and fence zones against design procedures, attending factory tests and assisting
documents to verify that the devices are correctly integrators with pre-commissioning tests, he has
mounted and aimed. tested and commissioned security systems on many
complex security projects in electric utility markets.
Labeling — The accuracy of labels on equipment in
cabinets and field devices should be verified through
comparisons to design documents, security workstation
02372-COM-0817
Tim Farrar
Chief Engineer-Protection and Controls
E: TFarrar@trcsolutions.com
P: (207) 621-7027
Page 266 of 284
Introduction
The protection system design for a typical substation involves many interrelated drawings, calculations,
studies and development of specific protective relay settings. Their accuracy has long been recognized
as critical to the successful, safe and secure construction, commissioning and operation of a utility
substation. This technical paper identifies a process that will improve the quality of protection system
design and protective relay settings to reduce the potential for protection system misoperations.
Refocusing attention on the Quality Control (QC) of protection system design and relay settings
development can achieve that goal.
Protection System Misoperations Defined
The North American Electric Reliability Corporation (NERC) is the regulatory authority whose mission is
to assure the reliable operation of the Bulk Electric System (BES) in North America. Under NERC’s
currently effective mandatory standard PRC-004-5(i) 1 effective April 1, 2017, utilities are obligated to
review all protection system operations in order to identify, report and correct those operations that
qualify as misoperations.
NERC has defined the term misoperation 2 as the failure of a Composite Protection 3 System to operate as
intended for protection purposes and lists six categories as described in Table 1 below.
The following is a general definition of misoperations:
• Any failure of a Protection System element to operate within the specified time when a fault
or abnormal condition occurs within a zone of protection.
• Any operation for a fault not within a zone of protection (other than operation as backup
protection for a fault in an adjacent zone that is not cleared within a specified time for the
protection for that zone).
• Any unintentional Protection System operation when no fault or other abnormal condition has
occurred unrelated to on-site maintenance and testing activity.
This definition is based on the IEEE Power System Relaying Subcommittee (PSRC) Working Group I3
“Transmission Protective Relay System Performance Measuring Methodology.”
1
The misoperations reporting standard PRC-004-5(i) may be found at this link:
http://www.nerc.com/_layouts/PrintStandard.aspx?standardnumber=PRC-004-
5(i)&title=Protection%20System%20Misoperation%20Identification%20and%20Correction&jurisdiction=United%2
0States
2
See the NERC Glossary of Terms for the full definition of the term misoperation:
http://www.nerc.com/pa/Stand/Glossary%20of%20Terms/Glossary_of_Terms.pdf
3
Composite Protection System - The total complement of Protection System(s) that function collectively to protect
an Element. Backup protection provided by a different Element’s Protection System(s) is excluded.
4
NERC’s Annual State of Reliability Reports may be found at this link:
http://www.nerc.com/pa/RAPA/PA/Pages/default.aspx
5
NERC Reports on outages due to Failed Protection systems through its portal at:
http://www.nerc.com/pa/RAPA/ri/Pages/AutomaticTransOutagesInitiatedbyFailedProtSysHErr.aspx
6
NERC MIDAS information may be viewed at this link: http://www.nerc.com/pa/RAPA/Pages/Misoperations.aspx
Failure to Trip - A failure of a Composite Protection System to operate for a Fault condition
During Fault for which it is designed. The failure of a Protection System component is not a
Misoperation as long as the overall performance of the Composite Protection
System is correct.
Failure to Trip - A failure of a Composite Protection System to operate for a non-Fault
Other than Fault condition for which it is designed, such as a power swing, undervoltage,
overexcitation, or loss of excitation. The failure of a Protection System
component is not a Misoperation as long as the performance of the
Composite Protection System is correct.
Slow Trip - A Composite Protection System operation that is slower than intended for a
During Fault Fault condition if the duration of its operating time resulted in the operation
of at least one other Element’s Composite Protection System.
Slow Trip - A Composite Protection System operation that is slower than required for a
Other than Fault non-Fault condition, such as a power swing, undervoltage, overexcitation, or
loss of excitation, if the duration of its operating time resulted in the
operation of at least one other Element’s Composite Protection System.
Unnecessary Trip – An unnecessary Composite Protection System operation for a Fault condition
During Fault on another Element.
The NERC 2017 State of Reliability Report states that while protection system misoperations rate
continue to decline, misoperations as a risk to reliability remains a priority. Figure 1 shows the total
correct operations rate for NERC through the first three reporting quarters of 2016.
Figure 1: Correct Protection System Operations Rate; NERC 2017 State of Reliability Report
Figure 2: Four-Year Misoperation Rate by Region (Q4 2012 – Q3 2016); NERC 2017 State of Reliability
Report
Figure 3 shows misoperations by cause code as a percentage to the total. The most common causes of
misoperations have remained the same through the past few years with over 60% percent of
misoperations caused by settings/logic/design errors, communication failures, and relay malfunctions.
Wolfgang EYRICH
entegra eyrich + appel gmbh
Germany
SUMMARY
Electrical power systems including substations have to be newly built and expanded continuously because
of the growing demand for energy. Due to this, a new philosophy for substation planning is necessary - the
Building Information Modeling (BIM) philosophy. The transition from conventional design to Building
Information Modeling presents a similar challenge to the engineers as it did for the transition to Computer
Aided Design (CAD) in the past. The Power industry is looking for the product -with capabilities beyond
that of existing CAD solutions - that allows for BIM solutions. BIM means new investments in software
and different approaches to projects. BIM is a design process that relies on an intelligent model - as the
basis for design, simulation and collaboration at each stage of the project life span. It provides quality, quick
and accurate execution of a wide range of different tasks including modeling, simulations, coordination,
and documentation.
This paper describes using of BIM models to simulate the timeline of the construction progress. It describes
the BIM 4D method - how a smart 3D model from an intelligent substation BIM software can be used to
simulate and check the construction progress. With the BIM 4D method, it is possible to visualize different
phases of the construction and define the necessary tasks and their duration on a timeline. With this BIM
4D method, different types of substation projects like extension projects or upgrades of already existing
substations can be handled and simulated comfortably.
KEYWORDS
Substation design, BIM 4D, 3D, substation BIM, construction progress, simulation, visualization
2. BIM
BIM is a design process that relies on an intelligent model as the basis for design, simulation and
collaboration at each stage of the project life span. It provides quality, quick and accurate execution of a
wide range of tasks including modeling, simulations, coordination, and documentation. [1] The BIM
workflow and philosophy – in the substation design context - means using one integrated information data
model in all the disciplines that are involved in the process. This means all included team members of the
substation design department, the building modeling, the building service engineering, the steel
construction, the terrain modeling and the electrical design work team. [2]
The core idea of BIM systems is a 3D model. In additional to the 3D visualization of the model (see
Figure 1), the model contains also a lot of other information e.g. quantities of materials, object properties as
shown in Figure 2 as well as other information that is necessary for the project documentation and
construction. If changes are made to the objects in an intelligent 3D model, these changes are reflected in
every element of the design where this object is used. With this, the designers can work more collaboratively
in one model. [3]
The timeline itself defines all the tasks. A timeline, as is well known from Gantt charts, illustrates a project
schedule as well as the start and finish dates of the elements of a project. The construction timeline can
therefore be defined for new substation projects (greenfield projects) as well as for extensions or upgrades
of existing objects (brownfield projects) - (see an example for a substation construction timeline in Figure 3).
Using only the traditional Gantt chart is difficult and time consuming – since that would make it necessary
to develop alternative schedules and to identify challenges and find the best way to achieve the best
construction progress solutions.
However, to handle the challenge of the replacement of equipment in substations or extensions of the
substations the following points need to be considered: “1) Construction has to be done around energized
equipment; 2) Offline time needs to be kept at a minimum; 3) Increase in performance and 4) Integration of
new technologies in to the existing state. The fourth point especially refers to older substations whose
electrical components do not comply with today´s standards or are no longer manufactured. Usually, in this
case, all electrical components should be replaced.
In order to perform some of the before-mentioned actions it is necessary to define the time intervals in order
to define de-energized state. Also, it is necessary to define switching operations of circuit
breakers/disconnecting switches for the purpose of physical enabling of de-energized state on the part of the
substation of interest. So the challenge is always to define the best way to handle the project to save time
and costs.” [2]. For these challenges, BIM 4D offers the perfect method to simulate and analyze the critical
time intervals.
4. BIM 4D – SIMULATION
As described before, BIM 4D enables designers in different disciplines to visualize the progress of
construction activities throughout the lifetime of the project. This approach has the very high potential to
improve project management and shorten the delivery times of construction projects, as well as being
scalable to any project size. The process of creating 4D models (BIM 4D) is be simple using a previously
created BIM 3D model - as smart models with all the necessary objects and project information. Those
objects or groups of objects in a BIM 3D model must then be linked to the tasks (construction activities) in
the predefined timeline. A simplified schematic representation of the 4D simulation process is illustrated in
Figure 4.
With a substation BIM based software (e.g. like primtech 3d) the foundation of a BIM 4D simulation is set.
A specialized 3D software export offers an optimized BIM collaboration with Autodesk Navisworks by
transferring - within the export - the BIM 3D substation structure (part stations, bays, modules, assemblies,
project symbols, etc.). The structure of such a BIM 3D substation model is described in a previously
published IEEE paper, [3]. In addition to the structure, it is necessary to transfer all the object types as well.
To prepare such a BIM 3D model, a 3D export (including layer conversion if necessary) is carried out and
then imported into Autodesk Navisworks. The next step - in Autodesk Navisworks - is to import the
predefined project plan (or newly created) from e.g. Microsoft Project Management to the TimeLiner in
Navisworks and to define Search Sets for each object. The final step is to attach the Search Set to the tasks
from the project plan. A BIM 4D simulation picture of such a workflow is shown in Figure 5.
As mentioned, the following actions are needed for the creation of the BIM 4D substation construction
model (see Figure 4 and 5):
1) Creation and export of a BIM 3D model with linked information in a substation BIM software
2) Definition of a timeline with tasks (Gantt chart) (see Figure 3)
3) Import of BIM 3D model in the BIM 4D simulation software
4) Linking of the intelligent CAD objects or structures to certain tasks.
CONCLUSION
As mentioned, BIM 4D offers the substation project team the possibility to 1) simulate the substation
construction progress, 2) test different possibilities to construction phases, 3) keep the offline time needs at
a minimum - through simulation, 4) control the time management and logistic of the substation construction
progress.
To simulate BIM 4D, a smart BIM substation model with linked information is needed - as is created in
primtech 3d – as well as the definition of construction phases and tasks. Finally, the smart 3D substation
model with the timeline and tasks can then be simulated and visualized in a simulation software.
[1] C. Eastman, P. Teicholz, R. Sacks, K. Liston, “BIM Handbook, A Guide to Building Information,
Modeling for Owners, Managers, Designers, Engineers, and Contractors” (Ed. 2. Hoboken, N.J.: John Wiley
& Sons, 2011).
[2] S. Heuser, W. Eyrich, M. Kokorus, “Smart substation information model - challenges in the
development of a Building Information Modeling (BIM) based software for substation design”, (CIGRÉ
Mexico International Colloquium, Mexico City, Oct. 2016).
[3] M. Kokorus, W. Eyrich, R. Zacharias “Innovative Approach to the Substation Design using
Building Information Modeling (BIM) Technology” (IEEE/PES May 2016, pages 1ff).
BIOGRAPHY
Wolfgang Eyrich is managing director and head of the software development department of the company -
Entegra eyrich + appel gmbh. He founded Entegra in 1991 and since then has further built a profound
expertise in the field of substation design while working together with a wide range of customers including
utilities, manufacturers and engineering companies. Furthermore, he has registered several patents. His main
areas of research are substations, optimization of substation design processes and integrated calculations.
He has published several technical papers.