Sie sind auf Seite 1von 12

MAT TOC

Proceedings of OMAE03
22nd International Conference on Offshore Mechanics and Arctic Engineering
June 8-13, 2003, Cancun, MEXICO

OMAE2003-37429
DESIGN, APPLICATION AND INSTALLATION OF AN X100 PIPELINE
Alan Glover, Joe Zhou and David Horsley
TransCanada Pipelines Limited., Calgary, Alberta, Canada
Nobuhisa Suzuki, Shigeru Endo and Jun-ichiro Takehara
JFE/NKK Corporation, Tokyo, Japan

ABSTRACT
Traditional pipeline technology will be severely
challenged as design-operating pressures continue to rise and
gas field developments occur in more remote locations
including the arctic. Cost-effective solutions to these issues
can be found through innovative designs using new technology
and its implementation. Some of these designs have considered
the use of high-pressure natural gas pipelines resulting in the
development of high strength steel. In order to meet these
increases in pressure TransCanada and JFE/NKK have been
working extensively on the application of X100 (Grade 690)
linepipe and this has culminated in the construction and
installation of a X100 project in the fall of 2002. This paper
will discuss the development of the related research projects
that allowed the successful completion of the field project. The
topics will include the material properties and fracture control
plans for X100. In addition the approach to strain based design
for X100 will include the analysis for both the tensile strain
limits (weld mismatch consideration) and compressive strain
limits (i.e. buckling capacity). The development of the field
welding process will also be covered. The paper will discuss
the implications of using X100 from the perspective of the
successful field project and the application of a strain-based
design.

INTRODUCTION
Traditional pipeline technology will be severely challenged
as design-operating pressures continue to rise and gas field
developments occur in more remote locations including the
arctic. Cost-effective solutions to these challenges can be
found through innovative designs using new technology and its
implementation. Some of these designs have considered the
use of increasingly higher natural gas pipelines resulting in the
development as shown in Figure 1. Not only have pressures

changed (and are continuing to increase) but also the regions in


which gas developments occur are becoming more remote and
in more hostile environments. Discussions on development of
gas reserves in the North American arctic have been ongoing
since the 1970s, and many factors have slowed or delayed
those developments. One of the reasons that the pipelines have
not been built is because previous studies have always found
the cost too high to make the pipeline economically justified.
In recent years changes in pipeline design and construction
technology have the potential for reducing the cost of these
Northern pipelines substantially.
TransCanada PipeLines
(TCPL) and JFE/NKK Corporation have been working on
various pipeline technologies that will contribute towards these
reductions in costs. Potential arctic designs will also experience
large temperature profiles as well as lower gas operating
temperatures and higher secondary loading. In order to provide
innovative solutions to these issues TransCanada has been
utilizing many innovative approaches including the use of high
strength pipeline steels. In order to use these steels many
different approaches have had to be developed. The following
material section summarizes the application of Grade 483 and
Grade 550 on the TCPL system. In order to accommodate
higher operating pressures, however, higher-grade steels have
been developed. The application of Grade 690 will be
described on a project that was completed in the fall of 2002.
These developments will contribute to cost-effective designs
and construction for the North.

MATERIALS
The prime impetus for increasing pressure in a gas
pipeline system and related increases in material properties is
economics. On a large diameter pipeline project in North
America about 40% of the project cost is related to material, (in
a Northern project about 30% of the cost relates to material),

Copyright 2003 by ASME

and hence reducing material costs has a significant effect on


project costs. A typical relationship is shown in Figure 2,
which clearly illustrates the benefit of using higher strength
material and is the driving force for increasing strengths to
even higher values. The figure only illustrates the effect up to
Grade 690 for moderate pressure increases and utilizes Grade
483 as the base analysis. Note that at some of the intermediate
pressures that there is no benefit from the increasing strength
level. This is because as the pipe material strength increases
this type of relationship has to also take into account D/t
practicalities, fracture control and influence of local buckling
behaviour (which will be discussed in detail later). Full
advantage of the higher grades can only really be taking into
account once a threshold pressure and wall thickness is
achieved. These factors become particularly important in a
strain-based design and applications for the North. This effect
is illustrated in Figure 3, which schematically shows that the
benefits of higher strength materials (e.g. Grade 690 or X100)
do not become effective until higher pressures are obtained.
TransCanada introduced high strength steel (grade
483, X70) into pipeline applications in the early 1970s. At the
same time higher design factors were implemented i.e. 0.8
design factors in Class 1 locations. Since that time over 6300
kms of large diameter Grade 483 pipe has been installed in our
system resulting in considerable savings for our shippers
through reduced tonnage. In 1994 and 1995 TransCanada
introduced the full use of Grade 550 (X80) on a 30 km project
(a small project had been completed in 1991). Since 1995
approximately 460 km of Grade 550 has been installed on the
TransCanada system. Grade 550 was successfully utilized in
discontinuous permafrost in Northern Alberta. The application
of Grade 550 on our projects has represented a 12% reduction
in the material costs. Note that for the range of pressures that
have been used, full advantage of the pipe wall reduction can
be achieved.
Grade 550 continues to offer advantages as design
pressures increase but at the high pressures wall thickness
becomes a limitation and higher strength steels are required.
Since 1998 TransCanada and JFE/NKK Corporation have been
working on the development and application of Grade 690
(X100), primarily as a potential application for high-pressure
gas pipelines. These projects have included the understanding
of the behaviour of the pipe material, the joining technology
required, the application of the material in a strain-based design
approach, fracture behaviour of the material and construction
related issues and have been reported separately in this
conference. As a result of these programs a decision was made
in early 2002 to implement Grade 690 on one of the summer
construction projects. The details of the project are reported in
a separate section. The success of this project results from a
technology program that has included:
Pipe Material Properties: All aspects of the material
properties have been studied as they relate to ordering and
performance. The key issues are around manufacturing
processes and chemistry, the stress-strain behaviour, yield to

tensile ratios (and how to measure these properties), uniform


strain, hoop and longitudinal properties, long seam weld
properties and toughness requirements. All this work has been
carried out on trial pipes, typically in the NPS 30 and 36 range.
In addition work is ongoing in terms of the effect of these
parameters on costing trends. The pipe used for the main
project was Grade 690 with a diameter of 1219-mm (48) and a
wall thickness of 14.3-mm (0.56). Specific details are given
later.
Joining: The emphasis to date has been on developing
mainline mechanized girth welding procedures and manual tiein procedures. The joining technology has also focussed on
developing procedures that would meet our strain-based design
for frost heave and for severe winter service. Procedures have
been developed for mechanized welding using pulsed GMAW
using standard wires and various gas mixtures. A lowhydrogen vertical down manual metal arc procedure or a
vertical up flux cored procedure is available for tie-in welds.
Considerable work is still underway, and this relates primarily
to higher productivity process/procedure and tie-in procedures.
Recent developments include twin wire, twin torch
applications.
Inspection: Standard inspection procedures can be used on
Grade 690 production welds, primarily using mechanized
ultrasonic inspection.
Work was performed on defect
detection, flaw acceptance and velocity calibrations.
Bending: Collaborative work showed that no issues would be
expected on bending of Grade 690. This work extended the
analysis performed on Grade 550, and the effect of springback.
In addition analyses were performed by CRC Evans to show
that the existing 48-inch bending machines would be capable of
bending pipe up to at least 16-mm wall. The results show that
the bends would require additional pulls to obtain the same
bend as compared to Grade 550. Collaborative work with BP
on a limited bending trial on NPS 36 Grade 690, also
confirmed the additional pulls.
Fracture Behaviour: Two full-scale fracture tests have been
completed on Grade 690 material as part of a Group Sponsored
project funded by TransCanada, BP, Alliance and BG
International and led by Advantica Technologies with support
from the Japanese steel and pipe makers. The two tests were
performed on NPS 36 material with wall thickness of
nominally 13.2 mm and 14.8 mm, at test pressures of 13,600
kPa (1972 psi) and 18,000 kPa (2610 psi). Both tests were
successful and fracture arrest was as predicted. The prediction
of fracture arrest was based on a correlation with Charpy
toughness and a correction factor. TransCanada is currently
working on an improved way of predicting the toughness
required for arrest that will eliminate the need for arbitrary
correction factors for both Grade 550 and 690. Nonetheless the
present full-scale results were used to develop the requirements
for material specifications for the prevention of fracture
propagation. Additional work is also being carried out on the
gas decompression characteristics of high-pressure gases and
this will become an important issue for those designs. Work

Copyright 2003 by ASME

also continues on a collaborative program with the Europeans


that will further extend the database on full scale fracture tests
for Grade 690.
Strain-Based Design: Extensive work is being performed
on developing the tensile and compressive strain limits
applicable to a high pressure design using high strength steel.
While these will not be required for this initial Grade 690
project (it will be a stress-based design), elements of it will be
incorporated into the approach including the defect tolerance
approach. This work has concentrated on the specification of
weld/pipe properties to achieve satisfactory tensile strain limits,
and on the numerical analysis of the approach. The local
buckling/post-buckling program continues to define the
compressive strain limit, however emphasis is on the material
behaviour in the post-buckling mode. This particular aspect is
reported separately in this conference.
Canadian CSA Code Implementation: While all of the
research work was being performed TransCanada was active in
developing acceptance of Grade 690 into the relevant CSA
codes. This resulted in incorporation of Grade 690 into the
new edition of CSA Z245.1-2002, which was ultimately
published in the fall of 2002. TransCanada utilized this in its
application to the Alberta Energy Utilities Board for approval
to utilize Grade 690. Close cooperation with the regulatory and
Code bodies was an important step in the rapid implementation
of this technology.
Applicability of Grade 690: In designing the application of
high strength steels (particularly for the North) and where
benefits can be achieved it is necessary to balance the
requirements of wall thickness requirements for pressure,
fracture control and buckling resistance (frost heave and thaw
settlement). While the pressure is a simple relationship, the
fracture and buckling capacity are a complex series of
interrelationships. In some cases increased wall thickness may
be required to provide fracture arrest, however as the pressure
increases for high-pressure designs the driving force reduces
and the relationship changes. Buckling capacity is a complex
relationship between strength (and stress-strain behaviour),
moment capacity and D/t ratios. This becomes apparent at the
lower pressures where Grade 690 offers no advantage over the
conventional Grade 550, however at high-pressures the reverse
is true.

WESTPATH PROJECT AND INSTALLATION OF


GRADE 690
Early in February 2002 an internal decision was made to
implement Grade 690 on one of the summer expansion
projects. This was followed up by a presentation to the
Operations Committee with a recommendation to proceed with
the installation of 1 km of NPS 48 on the Westpath project.
The Westpath project (Figure 4) involves building new
segments of the TransCanada pipeline system in Alberta and
British Columbia to help meet long-term growth gas demand in
the western United States and California markets. The 2002

$115 million project entailed building more than 86 kilometers


of pipeline loop along our existing system. The pipeline
project consists of 64 kms of Grade 550 NPS 48 and 22 kms of
NPS 20. The project also involves the addition of one
compressor station in Alberta, and modifications to the Elko
and Moyie compressor stations in BC. Adding to the
complexity is the fact that the Alberta System is regulated
under the Alberta Energy and Utilities Board, whereas the BC
System fall under National Energy Board jurisdiction. As a
result, the approvals for Westpath involved both the AEUB and
the NEB.
The specific installation of Grade 690 took place on the
Alberta Mainline Loop # 2 (Saratoga Section) in Alberta which
consisted 20.9 kilometers (13.6 miles) of 1219-millimetre (48inch) pipeline Grade 550 and 1.0 kilometers of 1219 mm (48inch) Grade 690. The schematic for the Saratoga loop is shown
in Figure 5.
The pipe material was supplied by JFE/NKK and ordered
to the CSA Z245-02 requirements plus TransCanadas
additional internal specification. The internal specification
places a much tighter tolerance on the pipe requirements than
the CSA code. One of the prime objectives of the project was
to gain experience in the manufacturing and construction of
Grade 690 so that it could be applied to future high-pressure
projects. The specific Saratoga project only required a design
of NPS 48 Grade 550 with a wall thickness of 12-mm. In order
to meet the objectives of the project and to develop longer-term
requirements for high-pressure designs it was decided to utilize
NPS 48 Grade 690 with a wall thickness of 14.3-mm for the 1km section of Saratoga. The design requirements for the pipe
were therefore based on that premise. This requirement meant
that some rapid development was required at JFE/NKK
resulting in some slight modifications to the U and O procedure
and to some of the welding requirements. Nonetheless all of
the technical and delivery requirements were met.
One key aspect of the specification of the material was
agreement on the type of testing to be performed to verify the
material minimum specified yield strength in the hoop
direction. Traditionally pipe material has been qualified using
the flattened strap specimen. TransCanada, along with others
has been evaluating the behaviour of high strength materials
using the flattened strap, round bar and ring expansion tests.
The results of these tests have shown that the flattened strap
presents a misleading approach to pipeline design because of
the Bauschinger effect and that the round bar test is an effective
method to qualify the material. This effect is particularly
pronounced for steels with strength levels greater than Grade
550 (Figure 6). The challenge is that if using a flattened strap
to qualify then the actual material yield strength is much higher
and the Y/T ratio becomes very high and often will exceed
0.96. Based on these analyses it was agreed that the Grade 690
pipe material would be qualified on the round bar procedure.
The flattened strap results were, however, collected to add to
the database and part of the continuing effort to have code
acceptance of the approach. Additional tests were also

Copyright 2003 by ASME

specified for the longitudinal stress-strain behaviour. These


results were for information purposes only but form part of the
strain-based design for the tensile strain criteria (namely
overmatching of the weld yield to pipe yield). All of the results
and comparison with the specification are given in Tables 1,2
and 3.
The results of the chemical analyses show that the pipe
met the additional requirements of TCPL P-04, with a product
CE of 0.26, typical of the prior trials. The results of the tensile
properties (Table 2) show that the pipes met the Grade 690
requirements of both CSA and P-04 when qualified with the
round bar specimen and as required. The average yield and
ultimate were 763 MPa and 838 MPa respectively, with a Y/T
of 0.91 (note that the maximum Y/T was 0.95). As expected
the flattened strap results did not meet the requirements in
terms of yield of the CSA code, note as well that the Y/T of
these specimens is much lower, again as expected. Our results
are also in agreement with the results published in Figure 6.
The longitudinal properties of the pipe gave slightly lower
yield and ultimate, and this was a deliberate action to enable a
more efficient strain based design for the tensile strain limits
(see later section). The flattened strap transverse weld samples
all met the CSA and P-04 requirements.
The fracture toughness property requirements of the pipe
and weld were determined based on a fracture initiation and
propagation control plan. The fracture arrest properties were
based on correlations from the full-scale fracture tests and from
conventional models with a correction factor. All of the
fracture toughness properties (Table 3) met those requirements.
Note CSA Z245.1 only addresses nominal pipe body
toughness. CSA Z662 (design requirements), addresses the
requirements for fracture initiation and arrest design, and for
higher pressures and stresses requires a full engineering
analysis.
A key requirement for the construction and installation of
Grade 690 was the qualification of the various welding
procedures. For the mainline this consisted of mechanized gas
metal arc procedures and for the tie-ins manual metal arc
procedures. The summary of the procedures is as follows:
1) Mechanized Gas Metal Arc Welding (GMAW) with a
vertical down welding progression were used for all mainline
welds as follows
Internal root beads shall be completed using short circuit
metal transfer with 75% Ar - 25% CO2 shielding gas
mixture and 0.9 mm Thyssen K-Nova wire
External weld passes shall be completed using pulsed gas
metal arc welding with a 85 %Ar - 15%CO2 shielding gas
mixture and 1.0 mm Oerlikon Carbofil NiMo-1 wire.
External cap pass shall be completed using short circuit
metal arc welding with a 85 %Ar - 15%CO2 shielding gas
mixture and 1.0 mm Oerlikon Carbofil NiMo-1 wire
100 C minimum preheat shall be maintained throughout.
2) Tie-in welds were completed using the shielded metal arc
welding (SMAW) process with a vertical down welding
progression as follows:

Root beads shall be completed with E5510-G (E8010-G),


minimum preheat 100 C maintained throughout
Hot, fill and cap passes shall be completed with 4.0 mm
Bohler BVD 110 (E11018-G)
Contractor shall ensure that there is no pipe movement until
after completion of the hot pass and there shall be a 24 hour
delay prior to inspection for all shielded metal arc welds
All of the welding procedures were qualified by both the
contractor and by TransCanada to meet the relevant CSA codes
and to be used for both workmanship and alternative
acceptance criteria according to Appendix K of CSA Z662-99.
Typical results from the procedure qualifications gave the
mechanized girth weld with average yield strengths of 698
MPa and ultimate strength of 815 MPa. The respective cross
weld tensile tests results all failed in the pipe material and gave
corresponding pipe longitudinal properties of yield strength
675 MPa and ultimate strength 811 MPa. Note these
longitudinal properties are slightly higher than those reported
for the pipe qualification in Table 2 (623 MPa yield and 801
MPa ultimate), however that is not unusual when performing
cross weld tests. In either case however the girth weld
properties overmatched those of the pipe longitudinal
properties and that was one of the main criteria. Additionally
prior to the commencement of the project detailed working
sessions were held with the contractor and for the welders the
welding procedures. This required that an extra welder training
school was set up immediately prior to kick off to re-train the
welders to utilize the pulsed gas metal arc procedures. This
was necessary in this particular case because the welders had
been on the overall Westpath project all summer constructing
the Grade 550 using mechanized short circuit gas metal arc
procedures. The change over to the pulsed procedures required
some additional training and also requalification. Views of the
internal and external welding are given in Figures 7 and 8.
Another potential issue with Grade 690 could have been
the field bending. As described earlier some preliminary trials
had been performed on NPS 36 and also calculations to show
that the bending could be performed using a standard CRC
bending machine with an internal mandrel. Nonetheless
because of the timing of the project and the delivery of the
pipe, it was not possible to do any pre-bending trials on the
NPS 48 Grade 690 material. Even so the field bending went
extremely well. No problems were experienced with the
bending, no coating issues arose, the pull times were similar to
the Grade 550 project, and slightly shorter pulls were used to
compensate for the additional springback. Overall bends of 1
degree per pipe diameter were easily achieved. The layout of
the pipe for the field project is shown in Figure 9.
Final field installation of Grade 690 took place in late
September 2002. After successfully training of the welders all
of the welding was completed over a 2-day period. The pipes
used for the project were all approximately 12 m in length and
no double jointing was performed. The pipes were left as

Copyright 2003 by ASME

single joints to permit the maximum number of welds to be


completed for the relatively short project. All of the pipes were
coated using standard fusion bond epoxy coating, with the
normal cut back to allow full ultrasonic inspection (Figure 10).
All of the field welding and inspection proceeded as planned.
Some lack of fusion defects were experienced, however, these
were all related to ongoing welder training as opposed to
welding process. Weld repair rates were similar to our other
start up mainline projects. Work is continuing to increase field
welding productivity using automated processes and these will
be introduced in the summer of 2003.
Complete hydrostatic testing of the line was performed in
early October and the line was placed in service November 1st
2002. Final meetings have been held with the regulatory
bodies to complete the information feedback loop in terms of
performance of the Grade 690 material. The pipe material is
now considered as acceptable for use in high pressure and
strain based designs. The following section describes the
approach that will be utilized on our strain-based design.

APPLICATION OF GRADE 690 IN A STRAIN BASED


DESIGN
Strain-Based Design
Strain based design applies to a subset of the limit states
where displacement-controlled loads dominate the pipeline
response. This would be typical for a pipeline design in the
arctic where secondary loads from ground movement would
dominate. In its application a procedure is also developed to
establish safety factors.
Note limit states are typically
classified into one of several broad categories, such as safety,
operability, and serviceability. Design criteria can then be
applied for example in terms of the external loading or
displacement conditions giving rise to the occurrence of each
limit state.
Load Events
The normal operating loads are determined from the
project Maximum Allowable Pressure and include any
contribution from the temperature differential between the
operating and installation conditions. During installation, the
pipeline is subjected to temporary installation loads such as
bending during laying and lowering. During operation and
dependent on the surrounding conditions, the pipeline may be
subjected to various external loads such as fault displacement,
ground movement, frost heave, thaw settlement and others. All
these loads and the design checks for them need to be described.
Dependent on the nature of a load event, it can be classified as
a load-controlled event or a displacement-controlled event. In
a load-controlled event, the magnitude of the load is
independent from the displacement and deformation of the
structure that the load applies to. Typical examples of loadcontrolled loads include self-weight, internal pressure, and the
constant external loads (forces) applied to the structure. A

load-controlled load is often described in terms of the direction


and magnitude of the applied force. In a displacementcontrolled event, the magnitude of the load applied to the
structure is dependent on the displacement and deformation of
the structure. Typical examples of displacement-controlled load
events are thermal expansion, frost heave, and imposed
displacements.
The significant difference between the load-controlled and
displacement-controlled events is that the structural responses
are fundamentally different beyond the peak load, as
conceptually shown in Figure 11. In the figure a steel bar
loaded with a pulling force P is shown in (a) as an example of a
system with a load-controlled load. Similarly, the same steel
bar is loaded with tensile displacement in (b) as an example
of a system with a displacement-controlled load. If the applied
load P and the imposed displacement increase consistently,
the structural responses of the load-controlled system (a) and
the displacement-controlled system (b) are identical as shown
in (c) until the ultimate stress is reached. At the ultimate stress
then a steel bar subjected to a load-controlled load fails in the
form of rupture.
Whereas a steel bar subjected to a
displacement-controlled load maintains its integrity and the
deformation process afterward remains stable and controlled
until the failure strain is reached.
For a structure with more complicated failure mechanisms,
such as local buckling, the fundamental difference of the
structural responses to a load-controlled event and a
displacement-controlled event remains the same while the peak
load (ultimate stress in above example) may be established
differently according to their specific failure mechanisms.
Similarly, the strength and deformation capacity governs the
resistances of a structure to load-and displacement-controlled
loads, respectively. Consequently, design criteria are either
strength-based or strain-based for load- or displacementcontrolled systems, respectively.
The applicable limit states for a particular pipeline are
determined based on potential and practical failure modes
relevant to the structure under the specific condition. Two key
limit states are determined to be failure of the weldments
(tensile limit) and piping local buckling (compressive limit).
Tensile Limit State Approach
Traditionally weld metal defects in pipeline girth welds
were accepted using workmanship based acceptance criteria.
These criteria have varied from country to country, however,
they have all been shown to be safe and conservative. This is
because the criteria were originally based on what was
considered to be a good level of welding workmanship and
have evolved over many years. While these criteria have been
shown to be safe, they do not have any basis in terms of design
of the pipeline, the loading on the pipeline, nor the materials
utilized in the construction of the pipeline.
In order to take advantage of changing welding
technology and cost-effective design solutions, many countries
have developed alternative methods for defect acceptance

Copyright 2003 by ASME

standards [CSA 1999 App K, AS-2885 1995, and API 1104


App A]. These alternative methods are all based on a fracture
mechanics approach and the acceptance criteria were developed
using relationships between defect size, material toughness and
stress conditions. Many of the standards have utilized
primarily a fracture and plastic collapse analysis, either
separately or combined within a failure assessment diagram.
These approaches have been applied for many years within
Canada [Glover 1981 and 1986], and recently within the
United States [Oil and Gas 1999]; all of these approaches,
however, were stress-based. The choice of a stress-based
assessment technique is most appropriate under load
controlled situations where the applied stress is low enough
that both the weld metal and pipe remain within their elastic
stress-strain response limits. When a pipeline is subjected to
displacement controlled loading, the applied strain may
exceed yield strain. For non-elastic longitudinal deformations,
the traditional stress based design criteria are inappropriate to
establish the relationship between defect tolerance and global
(pipe and girth weld) plastic straining capacity.
A typical strain-based design includes setting limits on
overall strain. The approach requires the understanding of the
collapse approach, and understanding the relationship between
the pipe and the weld properties, as well as defining a failure
criterion. This approach can be utilized when sufficient
information is known about the basic pipe and weld properties
and the original acceptance criteria utilized. TransCanada
Pipelines has been utilizing a strain-based approach for design
for several years and it is based on defining the tensile and
compressive strain limits. Generally speaking it has been
shown that the tensile strain limit can be increased if the weld
metal yield strength overmatches the pipe yield strength and
adequate toughness is achieved in the weld. In overmatching
welds the strain preferentially develops in the pipe material,
thereby shielding the weld from large plastic strains. If the
weld metal yield strength undermatches the pipe then increased
toughness in the weld is required to provide increased
resistance to the high strains that will develop in the weld.
Nonetheless, in undermatching circumstances, the limit on
tensile strain is reduced. If the weld undermatches by too much,
net section yielding will always control failure and the
allowable strain limit will be severely restricted. This approach
has been validated through a series of wide plate tests and finite
element modelling [Denys and Glover, 1994, Minami et al,
1995]
The results from the FEA and wide plate tests demonstrate that:
An acceptance criteria (allowable strain for a given defect
size) can be developed based on measured strain and
validated through full-scale tests.
Observation of many test results shows :
1. Failure strain> 0.5% results in Gross Section Yielding
2. Failure strain < 0.5% results in Net Section Yielding or
Fracture
3. Failure strain is reduced at high pipe Yield/Tensile ratios

4. Failure strain is reduced as the undermatching level of


the weld to pipe yield strength increases.
Hence by determining the distribution of properties within
the pipe and weld one is able to develop an approach for the
acceptable strain limit.
This overall approach includes
consideration of mismatch and Y/T effects on the tensile strain
limits. Previous work [Horsley et al 1997] has shown that weld
metal failure is expected to be controlled by gross section
yielding provided that the weld metal matches or overmatches
the pipe properties, and adequate toughness is achieved. In
addition if the weld undermatches the pipe slightly, then gross
section yielding can be obtained if the weld has reasonable
toughness, the defect size is not too large, and some
reinforcement is present.
This overall approach has been validated using a series of
wide plate tests and incorporated into some standardized
approaches (e.g. EPRG guidelines for defect acceptance,
Hopkins and Denys 1993). In addition FEA modelling
performed also showed that higher strains could be anticipated
for shorter defects. The limitation with the modelling approach
was that a failure criteria needs to be defined, which is
currently based on flow stress and this tends to underpredict the
behaviour for overmatched welds. The wide plate database
also showed the influence of the undermatching and
overmatching on the failure strains, as well as the effect of
yield to tensile ratio and the influence of defect size. This
overall approach has been confirmed using a series of project
specific wide plate tests. The results clearly showed that higher
strains to failure are obtained for small defects for the
overmatched welds compared to the equivalent undermatched
welds. The results also confirm that the premise of using an
overmatched weld with small defects allows higher strains to
failure to be achieved. Using this approach, criteria can be
established that will relate the failure strain to the tolerable
defect size. These criteria will be linked to the pipe and weld
properties and the inspection requirement for the weld. A
balance will then be achieved between the tensile strain limit
and the compressive strain limit for the specific loading
scenario. This approach has several implicit safety factors
already included in the analysis. Although these implicit safety
factors are important it is also recommended that a safety factor
be defined. To date the approach outlined in the BS 8010: Part
3: 1993, Code of Practice for Pipelines subsea: design,
construction and installation has been used. In this approach
the standard states that for situations where the displacements
of the pipe are bounded then the maximum allowable strain
should include a safety factor of 1.5.
Compressive Strain Limit
Local buckling of Pipe Section
When the pipeline is subjected to compressive loads
and/or bending moments, all or a portion of the pipe crosssection is experiencing compressive stresses and strains. When
the maximum compressive strain reaches a critical level, local

Copyright 2003 by ASME

buckling/wrinkling initiates in the pipe wall.


As the
compressive strains increase, the wrinkle continues to develop
which induces significant local deformation and as a result the
capacity to carry compressive loads is greatly reduced in the
buckled areas. If the primary load is load-controlled, local
buckling may immediately lead to significant cross-sectional
deformation and/or material failure (loss of pressure
containment). In the case of displacement-controlled loads, the
development of local buckling increases the localization of the
deformation at the wrinkle and accelerates the accumulation of
the strains. However, the pipe typically has significant
remaining deformation capacity beyond the onset
(initialization) of the local buckling before a true failure
condition occurs.
Extensive research has been conducted on the subject of
local buckling, wrinkling and post-buckling behaviour of pipe
in the last three decades [Bouwkamp and Stephen, 1973;
Gellin, 1980; Gresnigt, 1986; Lara, 1987; Mohareb, et. al.,
1993; Zimmerman, et. al., 1994; Zhou and Murray, 1995;
Yoosef-Ghodsi, et. al., 1995; Dorey, et. al., 1999; Das, et. al.,
2000]. The experimental and analytical studies have led to indepth understanding on the initiation and development of local
buckling and wrinkling. In general for common modern steel
pipes, the initiation of local buckling was found to be primarily
dependent on the D/t (pipe diameter to thickness) ratio and the
internal pressure, and to a lesser degree on factors including
pipe material properties, applied load combination, soil
restraints, initial geometric imperfection, and residual stress.
The initiation of local buckling is commonly represented
by the local buckling strain which is often defined as the total
compressive strain corresponding to the peak load in a loaddisplacement (e.g. moment-curvature) curve. The D/t ratio has
significant influence on the magnitude of the local buckling
strain. Increase in D/t ratio tends to reduce the local buckling.
The internal pressure dictates the local buckling modes. At
zero or a very low level of internal pressure, a typical local
buckling mode is shown in Figure 12a which is often referred
to as the diamond mode. At an intermediate to high level of
internal pressure, a typical local buckling mode is shown in
Figure 12b which is often referred to as the bulging mode.
Because of the different modes, the level of internal pressure
substantially influences the magnitude of the local buckling
strain. The local buckling strain increases as the internal
pressure level increases.
Based on numerous test data, a number of predictive
empirical equations have been developed to predict the local
buckling strain [Gellin, 1980; Gresnigt, 1986; BS8010, 1993;
Zimmerman, et. al., 1994; CSA, 1999]. Among all the
equations, the majority of them are intended for pipe under
pure bending, in other words, the effect of internal pressure is
not included. Overall equations developed by Gresnigt (1986),
Zimmerman et. al. (1994) and Dorey et. al. (2001) are more
appropriate for large size pipes subject to complex loading
conditions.

Post-Buckling of Pipe Section


For a pipeline subjected to load-controlled loads, the
initiation of local buckling (or the peak load) is considered to
be representative of the ultimate failure condition that is
defined as loss of containment or collapse of pipe cross section.
This is because the structural response in the post-buckling
regime is unstable and uncontrolled and the unstable process
eventually leads to failure. For a pipeline subjected to
displacement-controlled loads, the initiation of local buckling is
no longer a failure condition because of the inherent stability in
the displacement-controlled loading process in the postbuckling regime. It has been repeatedly demonstrated by fullscale experiments [Gresnigt, 1986; Zimmerman, et. al. 1994;
Dorey, et. al., 1999; Das, et. al., 2000] that pipes have
tremendous deformation capacity beyond the initiation of local
buckling. A recent test program, has studied various load
combinations (internal pressure, axial load/displacement,
bending deformation), and pipe specifications. It is concluded
from the test program that the pipe materials are highly ductile
and do not fail (loss of containment) when they are subjected to
monotonically increasing compressive strain and before the
wrinkle is fully developed and the faces are in contact.
Compressive Strain Limits and Safety Factors
For pipelines in permafrost areas, the loads that could
potentially induce excessive stresses and strains include frost
heave, thaw settlement, slope movement and temperature
differential. It is apparent, based on the mechanism for these
load events, that these loads are displacement-controlled loads.
For displacement controlled loads, the initiation of local
buckling is not a failure condition for a pipeline. It is, however,
recommended that compressive strain limits be established
based on the local buckling strain with a safety factor of 1.0.
Note this represents the initiation of local buckling and is
defined at the peak load point on a load-displacement curve, for
the following reasons. First of all, once a wrinkle initiates, it
develops relatively quickly because of the reduced load
carrying capacity in the wrinkled section. A developed wrinkle
has significant local deformation of the pipe wall and pipe
cross-section that may affect the functionality of the pipeline.
In addition, the current industrial practices of pipe-soil
interaction analysis are based on models where pipe is
simulated by a series of beam elements. The beam elements are
not able to properly simulate the local pipe wall and cross
sectional deformations, and consequently, a compressive strain
limit beyond the local buckling strain would be difficult to
implement in the design process.
For any given pipeline with defined pipe specifications,
material properties, manufacturing process and associated
operating conditions, then the compressive strain limit can be
established by utilizing an empirical equation. Supplemental
validation of the predicted compressive strain limit can be
provided by full scale local buckling tests when project specific
pipes are available. Since the compressive strain limit is
established based on maintaining the pipeline functionality and
any limitation on the current practice in pipe-soil interaction

Copyright 2003 by ASME

analyses, rather than the pipeline failure condition, hence the


remaining safety margin associated with this limit is
significantly larger than what would normally be required.

FROST HEAVE AND THAW SETTLEMENT


Permafrost presents a unique challenge to the design and
construction of a northern pipeline. Design for frost heave and
thaw settlement in permafrost can be one of the most
challenging aspects of a northern pipeline. To adequately
design for permafrost, balanced efforts on various aspects in
the following are required to achieve optimum overall
performance and effectiveness:
Establish appropriate pipeline operating strategies and
conditions, especially the temperature profile along the
ROW and the temperature cycle over time. This task is
commonly accomplished through a series of hydraulic
simulations of the pipeline system, including configuration
of surface facilities such as compressors and chillers.
Predict frost heave and thaw settlement over the entire
design life of the pipeline based on a thorough
understanding of soil, climate, and pipeline operating
conditions. This task is often accomplished through a
series of geothermal analysis for representative sections
and unique sections of the ROW.
Predict pipeline response resulting from the predicted frost
heave and thaw settlement, and comparing the predicted
maximum strains to the strain-based design criteria
discussed in the previous sections. This task is normally
accomplished through a series of pipeline structural
analyses, which include two main components: a definition
of the characteristics of soil springs that represents the soilto-pipe load transfer mechanism, and a pipe-soil
interaction analysis to determine the pipeline response to
the imposed frost heave and thaw settlement.
While the design process is divided into three aspects, it is
important to recognize the linkages and coupling effects among
them. It is obvious that operating conditions have a direct and
significant effect on the ground thermal state. The ground
thermal state also has an effect on the pipeline temperature
profile with varying degrees of sensitivity. Similarly, the
magnitude and rate of frost heave and thaw settlement are
dependent on not only the ground thermal state but also the
pipeline response resulting from the imposed frost heave and
thaw settlement. The interactive nature of the design process
for frost heave and thaw settlement requires integrated design
tools and process that are able to:
Capture the coupling effect and determine reliably and
accurately the frost heave, thaw settlement and pipeline
response;
Optimize the pipeline design, facility configuration and
operating strategies based on the overall system
performance and cost effectiveness;

Optimize the life cycle cost by balancing initial capital cost


and the ongoing operating and maintenance cost.
The overall design process and frost heave and thaw settlement
that TransCanada has been developing is conceptually
illustrated in Figure 13.

SUMMARY
Challenges in pipeline applications will continue to be met
though the application of innovative technologies and the use
of high strength pipeline steels. It has been shown that these
technologies can provide safe and reliable systems whilst at the
same time enabling cost-effective solutions. TransCanada
PipeLines has been at the forefront of some of these changes
and continues to seek alternative solutions that will drive down
the cost of major projects. The use of higher strength pipeline
materials, alternative pipeline materials, innovative designs
including strain and reliability-based approaches, structural
integrity solutions and alternative construction technologies are
all contributing to the ability to meet these challenges. Some of
these challenges can be met through the use of high strength
pipeline technology. TransCanada together with JFE/NKK
have successfully developed and installed Grade 690 (X100) as
part of the Westpath project in the fall of 2002. The use of
Grade 690 for these challenging environments has now been
successfully demonstrated and will now be part of costeffective solutions for high-pressure pipeline systems.

REFERENCES
American Petroleum Institute, (1994), API Standard 1104-18th
Edition, App A, May.
Australian Standard, (1995), AS-2885.2: 1995, Part 2:
Welding
Bouwkamp, J.G. and Stephen, R.M., (1973), Large Diameter
Pipe Under Combined Loading, ASCE, Transportation
Engineering Journal, Vol. 99, No. TE3, pp. 521-536.
BS8010, (1993), Code of Practice for Pipelines, Part 3,
Pipelines
Subsea:
Design,
Construction
and
Installation, British Standard Institute.
CSA (1999), CSA Z-662-99, Oil and Gas Pipeline Systems
Appendix K
Das, S., Cheng, J.J.R., Murray, D.W., Wilkie, S.A., and Zhou,
Z.J., (2000), Laboratory Study of Local Buckling,
Wrinkling Development, and Strain for NPS12 Line
Pipe, Proceeding of International Pipeline Conference,
October.
Denys R and Glover A.G., (1994), Conf. Mismatching of
welds, ESIS 17, London
Dorey, A.B., Murray, D.W., Cheng, J.J.R., Grondin, G.Y. and
Zhou, Z.J., (1999), Testing and Experimental Results
for NPS 30 Line Pipe Under Combined Loads,
Proceeding of the 18th OMAE Conference, Paper No.
OMAE99/PIPE-5022.

Copyright 2003 by ASME

Pressure kPa

25000
20000
15000
10000
5000
0
1900

1920

1940

1960

1980

2000

2020

Year

Figure 1 Pipeline system operating pressures by year

110
105

7000 kPa

100
% Cost
95

8000 kPa
9000 kPa

90

10000 kPa

85

12000 kPa

448

483

550

690

15000

SMYS

Figure 2 Effect of Pipe grade on overall project costs as a


function of pressure design using Grade 483 as a base

Wall thickness and pressure as a


function of D/t limitation
Wall thickness
factor

Dorey, A.B., Cheng, J.J.R., and Murray, D.W., (2001), Critical


Buckling Strains for Energy Pipelines, Structural
Engineering Report No. 237, Dept. of Civil and
Environmental Engineering, University of Alberta,
Edmonton, Alberta, Canada.
Gellin, S., (1980), The Plastic Buckling of Long Cylindrical
Shells Under Pure Bending, Int. Journal of Solids and
Structures, Vol. 16.
Glover, A.G., Coote, R.I., and Pick, R.J., (1981), "ECA of
pipeline girth welds," Int. Conf. on Fitness for Purpose
Validation of Welded Constructions, London,
November
Glover, A.G., Coote, R.I., and Pick, R.J., (1986), "Alternative
girth weld acceptance in the Canadian gas pipeline
code," 3rd Int. Conf. on Welding and Performance of
Pipelines, London, November
Gresnigt, A.M., (1986), Plastic Design of Buried Steel
Pipelines in Settlement Areas, Heron, Volume 31, No.
4.
Horsley D.J, Glover A.G. and Denys R. (1997), An assessment
technique for defects in under and overmatched pipeline
girth welds PRCi/EPRG, 11th Biennial Joint Technical
Meeting on Line Pipe Research, Arlington, Virginia.
Hopkins, P and Denys R., (1993), Background to the EPRGs
Girth Weld Limits for Transmission Pipelines,
EPRG/PRCi, 9th Biennial Meeting on Line Pipe
Research,
Knauf G. and Spiekout J. (2002), 3R International Special
Edition 13/2002
Lara, P.F., (1987), Revisiting the Failure Criteria of Buried
Pipelines, ASME, Petroleum Division (Publication),
PD. Vol. 6, pp. 143-154.
Minami F et al, (1995), Pipeline Technology Conference,
Ostend, Belgium,
Mohareb, M., Alexander, S.D.B., Kulak, G.L., and Murray,
D.W., (1993), Lab Testing of Line Pipe to Determine
Deformational Behaviour, Proc. 12th OMAE Conf,
Vol. V,
Oil and Gas Journal (1999), Article on the Alliance Pipeline
Yoosef-Ghodsi, N., Kulak, G.L., and Murray, D.W., (1995),
Some Test Results for Wrinkling of Girth-Welded Line
Pipe, Proceeding of 14th OMAE Conference, Vol. V,
Zhou, Z.J. and Murray, D.W., (1995), Analysis of PostBuckling Behaviour of Line Pipe Subjected to
Combined Loads, Int. Journal of Solids and Structures,
Vol. 32, No. 20.
Zimmerman, T.J.E., Stephens, M.J., DeGeer, D.D., and Chen,
Q., (1994), Compressive Strain Limits for Buried
Pipelines, Centre for Engineering Research (C-FER),
200 Karl Clark Road, Edmonton, Alberta, Canada.

120
100
80
60
40
20
0

X80
X100

1260

1440

1760

Pressure psi

Figure 3 Beneficial effect of increasing pipe grade as a


function of pressure but factoring in D/t limitations

Copyright 2003 by ASME

Table 1 Chemical analysis


Type of
Analysis

Heat and Product Analysis (weight %)


C

Si

Mn

Cu

Ni

Cr

Mo

Nb

V+
Nb

Ti

CE

Max.

Max.

Max.

Max.

Max.

NS

NS

NS

NS

Max.

Max.

NS

Max.

NS

Max.

0.26

0.50

2.00

0.030 0.035

0.11

0.11

0.11

0.40

TCPL P-04 & TA #2, Rev. 0 Max.


(Heat & product)
0.07

Max.

Max.

Max.

Max.

Max.

Max.

Max.

Max.

Max.

Max.

Max. 0.004 Max.

Max.

0.35

1.95

0.020 0.001

0.30

0.30

0.10

0.30

0.06

0.02

0.08

0.020 0.009 0.320

Actual (Average)

Ladle

0.06

0.10

1.87

0.009 0.001

0.27

0.14

0.03

0.22

0.05

0.00

0.05

0.009 0.005

0.28

Actual (Average

Product

0.05

0.09

1.87

0.009 0.001

0.28

0.13

0.03

0.21

0.05

0.00

0.04

0.008 0.005

0.26

CSA Z245.1-02
(Heat & product)

Table 2 Tensile properties


Spec &
Heat
No. For
Grade 690
Production

Pipe Body - Transverse


Flattened Strap Specimens
YS

TS

MPa
CSA
Z245.1-02

690
825

TCPL P-04 and


TA #2, Rev. 0

NS

Actual Average

684

Y/T

YS

TS

MPa

Ratio

MPa

760

Min

Max.

690

970

17

0.93

825

NS

NS

NS

690

27

Weld Transverse

Round Bar Specimens

Flattened Strap
Specimens

Round Bar Specimens

EL

846

Pipe body - Longitudinal

0.81

EL

Y/T

YS

TS

MPa

Ratio

MPa

760

Min

Max.

NS

970

11

0.93

760

Min

NS

NS

825

970

11

763

838

21

0.91

623

EL

Y/T

TS

MPa

Ratio

MPa

NS

NS

NS

760

Min

970

10

NS

NS

NS

Min

Min

760

10

811

12.4

801

22.3

0.78

EL

Table 3 Fracture toughness properties


Pipe Body, Weld, and Heat Affected Zone Toughness Transverse Specimens

Spec &
Heat
No. For
Grade 690
Production

Charpy Impact Tests @ -5C

Value
CSA
Z245.1-02
TCPL P-04 &
TA #2, Rev. 0

Results

Drop Weight Tear Tests @ -5C

Body
(any heat)

Body
(AHA)

Weld

H.A.Z.

Energy

Shear
(any heat)

Shear
(AHA)

(J)

(J)

(J)

(J)

(J)

(%)

(%)

NS

NS

NS

50

85

75

75

NS

85

90

40
140

210

Average

241

112

122

7781

100

Minimum

214

98

94

7059

100

All Heat
Average

241

100

10

Copyright 2003 by ASME

Figure 4 Overview of the Westpath project,


June to October 2002

First installation World wide


X100, 1000 m
NPS 48

Existing
Pipeline

Figure 7 Internal welding on Grade 690, standard short circuit

Standard Technology
X80 20900 m
NPS 48

Figure 5 Schematic of Saratoga Project


900

Rt0.5 - transverse

Rt0.5 on transverse flattened


rectangular specimen, MPa

800

700

UOE

X100
SWP

600
HFI

GRS 550, X80


500

400

x=y

flattened strip - round bar

HRS

X70
-

X60
X52

300
300
EPRG28VP3R

400

500

600

UOE manufactured pipe


HFI welded pipe
SWP spiral weld pipe
Hot rolled seamless pipe

700

800

900

Figure 8 External fill passes on Grade 690


using pulsed GMAW procedures

Rt0.5 on transverse round bar specimen, MPa

Figure 6 Comparison of the yield strength values measured on


flattened strap rectangular specimens with those
measured on transverse round bar specimens for
different pipe types
(from G. Knauf and J. Spiekout, 2002)

11

Copyright 2003 by ASME

Stress
EA

EA

EA

EA

ultimate
stress
yield
stress

Strain
P

yield
strain

failure
strain

(a)

(b)
(c)
Figure 11 Schematic of load and displacement controlled event

Figure 9 Pipe layout after all field bending had been completed

(a)
(b)
Figure 12 (a) Bulging mode buckling at high internal pressure
and (b) Diamond mode buckling at zero or low pressure

Hydraulic simulation to
predict temperature profile
over the design life

Establish tensile strain limit


based on material property,
welding and inspection

Geothermal analysis to
predict frost heave over
the design span at all
identified sites

Establish compressive
strain limit based on
material property, pipe
geometry and pressure

Structural (pipe-soil
interaction) analysis to
predict stress and strain

Design criteria based on


strain limits

Reliability based design methodology to ensure the target


reliability levels are met

Figure 10 Typical coating and marking for the Grade 690 pipe

Figure 13 Overall Design Approach for Frost Heave


and Thaw Settlement

12

Copyright 2003 by ASME

Das könnte Ihnen auch gefallen