Anda di halaman 1dari 10

OTe 6335

Design of Submarine Pipelines Against Upheaval Buckling


A.C. Palmer, Andrew Palmer & Assocs.; C.P. Ellinas, Advanced Mechanics & Engineering;
D.M. Richards, Lloyds Register; and J. Guijt, AlS Norske Shell

Copyright 1990, Offshore Technology Conference

This paper was presented at the 22nd Annual OTC in Houston, Texas, May 7-10, 1990.

This paper was selected for presentation by the OTC Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper,
as presented, have not been reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessaril'y reflect
any position of the Offshore Technology Conference or its officers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The
abstract should contain conspicuous acknowledgment of where and by whom the paper is presented.

ABSTRACT upheaval buckling incidents in the Danish


and Norwegian sectors, and by the cost of
This paper describes part of a protective measures such as rock dumping.
comprehensive joint-industry project on Shell International Petroleum Maatschappij
upheaval buckling. It develops a semi- (SIPM) initiated a comprehensive upheaval
empirical simplified design method and buckling research program in 1987. In its
detailed design methods based on a new final form, the research program included
numerical analysis, and illustrates their six phases:
application by examples. It assesses
alternative design strategies, and the
implications of strain-based design. 1 review of available models;
2 comparative analysis;
3 cover response, testing and modelling;
INTRODUCTION 4 numerical upheaval buckling model;
5 analysis of alternative concepts;
6 development of an upheaval buckling
When a pipeline is operated at a guideline.
temperature and pressure higher than
ambient, it will try to expand. If the
line is not free to expand, the pipe will The present paper Ls one of a series
develop an axial compressive force. I£ the describing the results of the programme.
force exerted by the pipe on the soil It concentrates on phases 5 and 6, which
exceeds the vertical restraint against were subcontracted by SIPM to a joint
uplift movement created by the pipe's venture of three UK engineering companies,
submerged weight, its bending stiffness, Andrew Palmer and Associates, Advanced
and the resistance of the soil cover, the Mechanics and Engineering, and Lloyds
pipe will tend to move upward, and Register.
considerable vertical displacements may
occur. The pipeline response might then be The paper is primarily concerned with the
unacceptable because of excessive vertical design aspects of upward buckling of buried
displacement or excessive plastic yield pipelines. Unburied pipelines can buckle
deformation. Upheaval buckling is hence a in a related but different mode, in which
failure mode that has to be taken into they snake sideways across the seabed.
account in the design of trenched and This mode too is important in practice, and
buried pipelines. can play a major part in relieving the
effects of axial force, but it is outside
The possibility of upheaval buckling is of the scope of the present paper.
increasing concern to the operators of
flowlines in the North Sea and elsewhere. The paper first briefly describes the
Their concern has been heightened by phenomenon, and puts forward a simplified
method for preliminary design. It goes on
to describe a conceptual design method
References and figures at end of paper based on the application of the UPBUCK

551
2 DESIGN OF SUBMARINE PIPELINES AGAINST UPHEAVAL BUCKLING OTC 6335

program, and applies the methodS to a An arbitrary pipeline profile can be


design example. The paper then considers defined by a height y and a horizontal
alternative strategies to resolve upheaval distance x. The pipeline is idealised as
problems, and examines in more detail one an elastic beam which carries an axial
of these strategies, a reduction in wall compressive force P and has flexural
thickness made possible by the replacement rigidi ty EI. It follows from elementary
of stress criteria on longitudinal stress beam-column theory that the downward force
with strain-based design criteria. w(x) per unit length required to maintain
the pipeline in equilibrium in this
position is
UPHEAVAL BUCKLING

The source of upheaval buckling is an


interaction between the axial compressive
force and overbend "hill" imperfections in and therefore depends on the profile shape
the pipeline profile. through the fourth and second derivatives.
Figure 1 illustrates a sequence of events Consider first a simple sinusoidal profile
which initiates buckling in a buried imperfection of height S and length L
pipeline. The pipeline is laid across an defined by
uneven seabed (Figure la), and later
trenched and buried (lb). The trenching
and burial operations modify the profile of
the foundation on which the pipe is y = 5 cos 2 (rrx/L) in -~L<x<~L ... (2)
resting, so that it is not precisely the
same as the original profile. Trenching The downward force required to maintain
may smooth the profile overbends, but may this profile is therefore
also introduce additional imperfections,
if, for instance, a lump of bottom soil
falls under the pipe. w(X) = (-8SEI(rr/L)4 + 2SP(rr/L)2)cos(2rrx/L)
.., (3)
When the pipe goes into operation, its
internal pressure and temperature are and has its largest numerical value at the
higher than when it was installed and crest of the imperfection, where
trenched, and the axial force becomes
compressive. The effective axial force in
a constrained pipeline has two components,
both of which contribute towards buckling w 25P(rr/L)2- 8SEI(rr/L)4 ...... (4)
(Palmer 1). The axial force in the wall is
the resultant of a compressive constrained
thermal expansion component and a tensile the downward force per unit length required
Poisson component, while in addition there to stabilise the pipeline at the crest of
is a compressive force component in the the profile imperfection.
contained fluid.
Equation (4) applies to a particular
On an overbend, the axial compressive force profile shape, but the specific shape only
reduces the upward reaction between the affects the coefficients and not the
foundation and the pipeline (Figure lc). general form of the equation. It can be
A further increase of operating temperature rewritten
and pressure may reduce the reaction to
zero. The pipe then lifts on the overbend,
moves towards the surface of the cover, and
may break out through the surface
(Figure ld).
a relationship between a dimensionless
maximum download parameter ~w defined by
SIMPLIFIED ANALYTICAL MODEL
~w wEI/SP2 ..................... (6)
The stability of the pipeline in its
initial position turns out to depend on the
local profile of the pipe in contact with and a dimensionless imperfection length ~L
its foundation, and on whether or not defined by
enough downward force is present to hold
the pipe in position. If the pipe does not
move, the governing factors are the ~L L ( P / E I ) ~ •••••••••••••••••.•• ( 7 )
constrained axial force and the pipe
flexural rigidity, and other parameters
have no effect. This suggests that a useful way of
representing the results of upheaval
buckling calculations is to plot ~w against

552
OTC 6335 PALMER, ELLINAS, RICHARDS, GUIJT 3

~L' The resulting plot is universal, and a design formula for the required download
forms a valuable summary, which can include for stability in the operating condition:
both the results of numerical calculations
and then observations from full-scale field
experience and tests. w [1.16 - 4.76 (EIwo/cS )~/P]P( cSwo/EI)l.:;
... (12)
All points that represent a single
imperfection shape will lie on a single where
curve on a ~w against ~L plot. Points that
represent different imperfection shapes Wo is the installation submerged weight;
will lie on different curves, and the EI is the flexural rigidity;
variability between curves will reflect the cS is the imperfection height;
effect of imperfection shape (the hardest P is the effective axial force in
parameter to determine reliably in operation.
practice). The form of equation 5 suggests
that the functional relationship between ~w This equation is used to derive the
and ~L is required value for download for preliminary
design. Preliminary design calculations
apply a spreadsheet, which compares the
~w C~L-4 + d~L-2 ••••••••••••••••• ( 8) required download determined from equation
(12) with the actual load that can be
mobilised, the sum of the pipeline's
where c and d are constants to be submerged weight and the uplift resistance
determined numerically. This is done by of the cover. The research carried out
plotting ~w~L 2 against ~L-2, in Figure 2, on into uplift resistance in the course of
which the points represent the results of this research programme is discussed in
numerical calculations with the UP BUCK detail in a separate paper2. In brief, the
program described below. The rightmost model recommended for design of buried
group of points represents imperfection pipelines is
profiles whose shape is an upheaval
foundation, that is, a foundation in the
shape of a pipeline supported by axial for cohesionless sand, silt and rock
force in a post-upheaval mode. The
remaining points correspond to "prop" q = HD(l + fH/D) ............... (13)
imperfection profiles, whose shape is that
taken up by a pipeline laid across a single
isolated hill in a horizontal profile. The for cohesive clay and silt cover
final relationship turns out to be
bilinear, with one pair of values c and d q cD min[3,H/D] .............. (14)
corresponding to small values of ~L and a
second pair to large values. where
An additional condition occurs when the q is the uplift resistance per unit
profile includes a very short imperfection, length of pipe;
so that the pipeline is only in contact H is the cover depth (from the top of
with the crest of the imperfection. The the pipe to the surface);
length ~L is then immaterial, and this D is the outside diameter of the pipe;
condition forms a cut-off for short is the submerged unit weight of the
imperfections. cover material;
c is the shear strength; and
The results are combined in Figure 3, a f is an uplift coefficient, determined
universal curve which can be used directly experimentally and taken as 0.5 for
for design. The three conditions that dense materials and 0.1 for loose
together give the design download required materials.
for stability are:

The reader is warned that this simplified


~L < 4 • 4 9, ~w = O. 0 6 4 6 .•..•..••.••.• ( 9 )
method is semi-empirical, and that on its
4 • 4 9 < ~L < 8 • 0 6, ~w 5 . 6 81 ~L 2 - 8 8 • 3 51 ~L4 •• ( 1 0 )
own it will not normally be adequate for
~L>8.06, ~w = 9.6/~L2 - 343/~L4 ... (ll)
design. It has been calibrated against
UPBUCK, but does not always yield
conservative results, especially if there
In most preliminary design contexts, the
is a possibility of plastic deformation of
designer can determine the maximum height
the pipe wall.
of a profile imperfection, but not its
length. An imperfection length can however
be estimated from an assumption that the
pipeline takes up a form dictated by the
interaction of its flexural stiffness and
its weight in the installed condition (not
in general the same as the weight in the
operating condition). Substi tuting this
into the third equation Ill). we arrive at
553
4 DESIGN OF SUBMARINE PIPELINES AGAINST UPHEAVAL BUCKLING OTC 6335

CONCEPTUAL AND DETAILED DESIGN taken as 0.3 m. The trench will be


backfilled with rock, which has a minimum
in-situ submerged unit weight of 8.5 kN/m 3
Conceptual and detailed design are based on and an uplift coefficient of 0.5. The
~he UP BUCK fini te-el~ment program described
assumed in-situ submerged unit weight is a
In a separate paper. The program takes low conservative value, and corresponds to
account of a number of factors not included a voids ratio of 0.87 and a rock fragment
in the simplified analytical procedure. density of 2650 kg/m 3 • The option of
These factors have a significant effect on trenching and rock-dumping, rather than
the response of the pipeline after it has rock-dumping alone, is selected because
begun to move in response to axial force analysis has shown that the saving in rock
and thence on its post buckling behaviou; cost more than outweighs the cost of
and the amplitude of the final trenching.
displacement. Among the factors are:
The wall thickness design arrives at a
nominal wall thickness of 11.1 rnrn. The
1 the finite axial stiffness of the installation submerged weight is 591.3 N/m
li~e, which determines how rapidly the (60.2 kg/m, 40.5 lb/ft), and the operating
aXlal force diminishes as the line submerged weight is 984.2 N/m.
moves upwardsi
The first step is preliminary design. It
2 the pipeline's longitudinal resistance calculates the factor of safety on pipeline
to movement through the soil which weight and cover uplift resistance taken
de~ermines how far the pipel.ine can together. This is the ratio between the
sllde towards a developing bucklei required value, calculated from equation
(12), and the actual value, the sum of the
3 the incremental flexural rigidity for operating submerged weight and the cover
large deflections, which determines uplift resistance taken from equation (13).
whether the pipeline's resistance to
b~nding decreases as large Figure 4 plots the calculated factor of
dlsp~acements occur: the rigidity may safety against the depth of cover. It can
be lnfluenced by plastic bending be seen that if the cover is 0.6 m or less,
deformation. the factor of safety is less than 1 that
~f the cover is 0.7 m the factor of ~afety
4 the finite stiffness of the lS 1.11, and that if the cover is 0.8 m the
foundation, which allows the fc:c~or is 1.32. This suggests that the
foundation to deflect under forces ffilnlmum cover should be at least 0.8 m,
applied by the pipe, and slightly sinc~ a 0.7 m cover leaves only a narrow
modifies the deflected form of the margln. Accordingly, a minimum cover of
pipe. 0.8 m is taken as the starting point for
UPBUCK calculations.

UP BUCK makes it possible for the user to A separate calculation determines the
calculate the response of the pipeline to extreme stresses in operation, at the
increasing operating pressure and imperfection and under the axial loads
temperature, and to follow the deformation induced by the operating temperature and
into the post-buckling range. pressure. Although the hoop stress is
r~latively low (183 N/rnrn2, 44 per cent of
Yleld), the longitudinal stress is high
DESIGN EXAMPLE and the von Mises equivalent stress is ave;
yield. This condition is allowed by many
codes, since the pipeline is continuously
The example is a 10-inch oil pipeline, with supported, but the high level of equivalent
s~res~ is an important warning that the
a maximum operating pressure of 15 MPa
~lpellne response to axial loading may be
(2175 psi) and a maximum operating
temperature of 80 deg C. The chosen steel lnfluenced by plasticity, and that it will
grade is API 5LiX60, which corresponds to be prudent to carry out detailed UP BUCK
a specified minimum yield stress (SMYS) of calculations using the plasticity option.
413.7 N/rnrn2. The external anti-corrosion
coating is 2 rnrn of polyethylene. The line Moving now to conceptual design, a series
is to be designed to DnV 1981 rules. of UP BUCK calculations are carried out. In
this phase of the analysis, the pipeline is
The line will be trenched, in order to treated as elastic.
protect it against fishing gear. A review
of existing survey data along a nearby The results of three UP BUCK calculations
pipeline route has established that the are summarised in Figure 5, which plots
seabed is relatively smooth, and that the movement against temperature increase for
maximum expected height of overbends will cover depths of 0.7, 0.8 and 0.9 m above
be 0.2 m. However, it is thought that the peak of the 0.3 m imperfection. If the
trenching might induce additional c,?ver. is 1. m (from the baseline), the
~lpellne ~eglns to lift (at the top of the
imperfections in the bottom profile, and
the design imperfection height is therefore lmperfectlon) at a temperature of about 70

554
OTC 6335 PALMER, ELL INAS, RI CHARDS, GUI J'l' 5

deg C, some 10 deg C short of the operating ALTERNATIVE DESIGN STRATEGIES


temperature. It becomes unstable and jumps
into a buckled configuration at 88.5 deg C.
This option leaves too little margin The simplest and most straightforward way
between the operating temperature and the of stabilising a pipeline against upheaval
temperature at which large movements might is to bury it. This option is often
occur, and is therefore rejected. difficult and expensive, and prompted a
search for alternatives. Phase 5 of the
If the cover is 1.1 m, movements begin at research programme examined some forty
about 80 deg C, and an unstable jump occurs alternative design strategies, evaluated
at 95 deg C. At the design maximum each one, and identified the most generally
operating temperature, the von Mises attractive and applicable.
equivalent stress is 83 per cent of yield,
which is within the maximum stress allowed Two obvious strategies are to reduce the
by the code. design operating temperature and pressure,
and to increase the submerged weight of the
If the cover is 1.2 m, movement does not pipeline. A reduction in operating
begin until the design temperature is temperature is generally impracticable, but
reached. The conclusion of the conceptual could be accomplished by adding a heat
design phase is therefore that a cover of exchanger to the system. It is not usually
1.1 m (from the baseline) is adequate, and practicable to resolve the problem by an
this value is chosen as the starting point increase in submerged weight, because too
for detailed design. much weight is required.

Continuing to detailed design, it has been The remaining strategies can be classified
shown earlier that the pipe is close to into groups. The most promising techniques
plastic yield, and that UPBUCK calculations are listed in table 1.
using the plasticity option are required.
The results show that movements initiate at The first group is based on reducing the
almost exactly the 80 deg C design driving force, the axial compressive force
temperature, but do not exceed 10 mm until in the pipeline. The most direct method is
the temperature has risen to 88 deg C. The to reduce the wall thickness of the line.
instability temperature is 97.3 deg C. This reduces the temperature component of
Figure 6 plots the movement of the pipe at the effective axial force, which is
the peak of the imperfection against the proportional to the wall thickness (and is
temperature, and shows that instability usually the largest component), and leaves
will be accompanied by very large the pressure component almost unchanged.
movements, in which the pipeline will break A reduction in wall thickness can be
above the surface of the cover. However, achieved by increasing the design factor to
there is an adequate temperature increase the highest allowable level, or by
marg in between the operating temperature increasing the grade of steel, or by
and the instability temperature. adopting strain-based design criteria, a
possibility examined below in section 7.
A cover of 1.1 fi, measured from the
baseline, is therefore selected. It A second method is to increase lay tension.
corresponds to 0.8 m cover at the peak of Residual lay tension balances part of the
the 0.3 m imperfection. It happens to compressive force induced by operating, and
coincide with the value reached by the therefore reduces the resultant force. A
simplified method, though this is not difficulty is that residual tension cannot
always the case. be measured directly, but must be
calculated from the lay conditions, and
This cover is only required on the that its continued presence in the line
overbends. If the critical overbends can depends on there being no lateral
be confidently identified from a survey, movements. A third alternative is to
rock need only be dumped on those preheat the line, and to allow it to move
overbends. It will usually be cheaper to to relax compressive forces induced by
devote resources to survey rather than to preheat: the application of this technique
continuous dumping. This design option was to the Sun Oil Glamis flowlines is
recently applied to the stabilisation of described in a separate paper5 •
the Tern and Eider pipelines 4 • If on the
other hand the s.urvey accuracy is not A second group of methods depends on making
sufficient to locate the critical a radical change in the structure of the
over bends , the whole length must be pipeline. One alternative is to replace
covered. one or more single lines by a closed bundle
supported on spacers in a carrier
(Palmer 6 ), or equivalently by a pipe-in-
pipe system in which an internal flowline
is supported in an outer pipe. The
internal lines in the bundle then develop
axial compressive forces in operation, but
those forces can be balanced by tensile
forces in the outer carrier, through end
555
6 DESIGN OF SUBMARINE PIPELINES AGAINST UPHEAVAL BUCKLING OTC 6335

bulkheads and possibly intermediate STRAIN-BASED DESIGN CRITERIA


bulkheads. The internal lines may bow
laterally on the spacers, but the movements
are controlled and the bending stresses In the past, the conventional approach to
induced may remain at an acceptable level. wall thickness design was generally
This effect can be further enhanced by governed by two stress requirements. The
making the internal bundle helical first requirement is that the hoop
(Duxburi) : this geometry increases the (circumferential) stress n~t. exc~e~ a
reduction in axial compressive force that defined fraction of the spec1f1ed m1n1mum
accompanies outward movements of the yield stress (SMYS). A second requirement
internal lines. A decision to opt for a also has to be satisfied, and takes account
bundle concept naturally has broader of longitudinal stress; the maximum value
implications that extend far beyond of an equivalent stress must not exceed a
upheaval buckling, which is only one of the specified fraction of the SMYS.. The
factors involved. specified fraction for the equ1valent
stress is generally higher than the
A second radical alternative is to replace specified fraction for the hoop stress.
a rigid steel pipeline with a flexible. The equivalent stress is defined eith~r as
Flexibles are subiect to upheaval buckling a von Mises equivalent stress (proport1onal
(Bournaze1 8 , Putot ,10), primarily because of to the square root of the second invariant
the pressure effect. The tendency to of the deviatoric stress tensor), or as the
buckle in service can be reduced by laying difference between the largest and smallest
or trenching under internal pressure, or by principal stresses.
modifying the internal structure away from
a "balanced" design, so as to produce a If this approach is adopted for pipelines
pipe which tends to contract axially when that operate at a high temperature, the
loaded by internal pressure. second requirement often determines the
wall thickness. This happens because the
A third group of methods is based on longitudinal stress in operation is large
stabilising the pipeline with rock, but and compressive, and because the
using the rock in a more efficient way, so temperature component of stress is
to gain an equally effective stabilising independent of the wall thickness. The
effect with less rock. The most effective need to reduce the equivalent stress below
of these methods is only to place rock on the allowable limit then forces down the
critical overbends (Locke 4 ), but this does allowable hoop stress, and this in turn
of course make it essential to be able to leads to a high wall thickness.
identify the overbends confidently, a
demanding survey task in deep water, or An alternative approach is better. In ~he
when the critical imperfection height is case of a continuously-supported
small. A second method is to place a constrained pipeline on or in the seabed,
geotextile over the pipe before the rock is exceedance of an allowable combined stress
placed. When the pipeline begins to lift, does not in itself correspond to a limit
the weight of the rock on the geotextile on state which in any way threatens the safety
either side of the pipe holds the of the line. If the combined stress should
geotextile down, and generates a tension reach yield, a limited amount of plastic
whose downward component adds to the uplift deformation will occur, but it can be shown
resistance of the rock above the pipeline. that the plastic components of longitudinal
At the same time, the geotextile pulls the and hoop strain are small by comparis<;>n
rock inward, and increases the horizontal with the elastic component of hoop stra1n
compressive stress in the rock above the (provided that the first requirement still
pipeline, and that in turn increases the limits the hoop stress, and that the design
shear resistance across potential shear factor is at customary level). The
surfaces above the pipe, further increas ing combined stress requirement can then be
the uplift resistance. Against this, on dropped, and replaced by a limit on strai~.
the other hand, the use of a geotextile in This is a more rational approach, and 1S
a subsea environment will require a allowed by several modern codes. The
comprehensive investigation of its long- allowable strain is set at a high level,
term stability against creep and structural often 0.01 or 0.02, but this level is
deterioration. almost never reached, except when the pipe
is deliberately plastically deformed in a
A final alternative is to not to place the construction operation such as J -tube pull.
rock continuously along the whole length of If the equivalent stress requirement is
the pipeline, but to place the rock in dropped and replaced by a limit on strain,
intermittent dumps. This possibility is the hoop stress condition determines the
discussed in an accompanying paperll. wall thickness. The resulting wall
thickness is then often significantly
The offshore pipeline industry is only now reduced. Figure 7 is an example: it refers
beginning to explore the many different to an 8-inch flowline with a design
concepts that can be used to secure temperature of 125 deg C and a design
effective and economical stabilisation, and pressure of 27 MPa (3910 psi), and plots
much more remains to be learned. the hoop stress, the longitudinal stress,
and the von Mises equivalent stress against
556
OTC 6335 PALMER, ELLINAS, RICHARDS, GUIJT 7

the wall thickness. If the equivalent Plasticity effects will therefore have a
stress requirement is adopted, and the significant influence on upheaval buckling
equi valent stress is limited to 0.96 of of pipelines designed according to strain
SMYS (as in the DnV rules), and the steel criteria. The solution will be sensitive
has an SMYS of 413.7 N/mm2 (corresponding to the details of the stress-strain
to API X60), the minimum wall thickness is relation for the steel, particularly to
12.8 mm. If on the other hand the strain-hardening. These effects are taken
equivalent stress requirement on into account in analysis by the UP BUCK
longitudinal stress is dropped, and the program.
hoop stress is limited to 0.72 of SMYS (as
in many codes), the minimum wall thickness
can be reduced to 9.6 mm. CONCLUSION

This reduction in wall thickness leads to


a substantial reduction in axial force in This paper summarises design methods based
operation, and to a corresponding reduction on the results of an extensive program of
in minimum cover required for stability. research into upheaval buckling.

Adoption of strain-based criteria does


however raise a number of wider questions, ACKNOWLEDGEMENT
which tend to reduce the advantages.
Figure 8 plots the stress path followed in
a sequence of laying, increase in The research described here was carried out
temperature, and increase in operating as part of a joint industry program led by
temperature, in a diagram whose axes are Shell International Petroleum Maatschappij
longitudinal stress and hoop stress. The and supported by the following companies:
ellipses represent successive yield loci
for a von Mises material with an isotropic Shell Internationale Petroleum Maatschappij
hardening rule. The heavy ellipse Maersk Olie og Gas A/S
represents the nominal yield locus, Nederlandse Aardolie Maatschappij BV
corresponding to the 0.5 per cent yield Shell UK Exploration and Production
locus, but yield begins before this locus A/S Norske Shell
is reached, and is represented Britoil Plc (BP Exploration)
schematically by the inner loci. UK Department of Energy
Marathon Oil UK Limited
If a strain criterion is adopted, the Elf Petroland BV
stress point S that represents the mean Occidental Petroleum (Caledonia)
longitudinal and hoop stresses during Saga Petroleum A/S
operation lies beyond the yield locus, even Sun Oil Britain Limited
if the hoop stress is less than 0.72 of Total Oil Marine plc
nominal yield. If the pipe then begins to
bend, the points representing the stresses The authors wish to record their gratitude
in the pipe in the "inside" and "outside" for permission to publish this paper.
of the curve follow the paths shown in
Figure 9b. On the inside of the curve, the
longitudinal stress becomes more REFERENCES
compressive, and the stress pOint moves to
the left. This takes it across additional 1 Palmer, A.C. and Baldry, J.A.S. Lateral
yield loci, and more yield occurs. On the buckling of axially-compressed
outside, the longitudinal stress becomes pipelines. Journal of Petroleum
more tensile, and the stress point moves Technology, ~, 1284-1284 (1974).
inward, so that the incremental response is
elastic. 2 Schaminee, P.E.L., Zorn, N.F., Schotman,
G.J.M. Soil response for pipeline
It follows that the incremental bending of upheaval buckling analysis: full-scale
a pipeline designed according_to strain laboratory tests and modelling.
criteria will always be accompanied by Proceedings, Twenty-second Annual
plastic deformation, however small the Offshore Technology Conference, Houston
bending is. The effect is to generate both (1990).
circumferential and longitudinal plastic
strain. The incremental increase in 3 Klever, F.J., van Helvoirt, L.C., and
bending moment is much smaller than it Sluyterman, A.C. A dedicated finite-
would be if the pipe were elastic, because element model for analysing upheaval
only part of the pipe is responding buckling response of submarine
elastically: the effect is to reduce the pipelines. Proceedings, Twenty-second
incremental flexural rigidity, which is one Annual Offshore Technology Conference,
of the parameters that enable the pipe to Houston (1990).
resist upheaval buckling. At the same
time, plastic deformations relieve some of 4 Locke, R.B. and Sheen, R. The Tern and
the axial compressive force in the pipe Eider pipelines. Proceedings, 1989
wall, and therefore reduce the driving European Seminar on Offshore Pipeline
force. Technology, Amsterdam (1989).
557
8 DESIGN OF SUBMARINE PIPELINES AGAINST UPHEAVAL BUCKLING OTC 6335

5 Craig, loG. and Nash, N.W. Upheaval 9 Putot, C.J .M. Localised buckling of
buckling: a practical solution using hot buried flexible pipelines. Proceedings,
water flushing technique. Proceedings, Twenty-first Annual Offshore Technology
Twenty-second Annual Offshore Technology Conference, Houston, OTC6155 (1989).
Conference, Houston (1990).
10 Putot, C.J .M. and Rigaud, J. Upheaval
6 Palmer, A.C. Are bundles expensive? buckling of buried pressurised
Proceedings, Pipelines for Marginal pipelines. Proceedings, European
Fields Conference, Aberdeen (1988). Seminar on Offshore Pipeline Technology,
Paris (1990).
7 Duxbury, P.G. and Hobbs, R.E.
Helically-laid pipeline bundles. 11 Ellinas, C.P., Supple, W.J. and
Proceedings, Eighth International Vastenholt, H. Prevention of upheaval
Conference on Offshore Mechanics and buckling of hot submarine pipelines by
Arctic Engineering, 2, 9-16 (1989). means of intermittent rock-dumping.
Proceedings, Twenty-second Annual
8 Bournazel, C. Flambage vertical des Offshore Technology Conference, Houston
conduites ensouillees. Revue de (1990).
l'Institut Francais du Petrole, ~, 212-
230 (1982).

I
II a as-laid

,
.. .
'. "

I 8 [
b trenched and buried
.~~
... . .
J
,I
",

c start-up

pipeline 'pUShes upwara.


against cover
• 2 r
0~1____~'____~,______~____~____~,__
o .81 .82 .83 .85
d "'h".~ .•.. ~-:-:
. . . .~
. .."". __
. . .._'_'.___ - -,
~:...-

Figure 2 Numerical correlation of UPBUCK results

Figure 1 Sequence of laying. trenching and upheaval

558
(]i
-'" j -l • -l

8. <)6
0<2,,-1.-19
=<I, \ 8 . 06
<:I,
1:. ~ <J. 06 -I 6

I .. - ~ :=-;
;2"

I .. - 9..:.: - 3':?
.!,
- =-3.
.!,

1.,
2

ratio l.S
. 06 <0 available/required
o downward force
o
. 0..l 'Q
o
S

.0: I c 0.5

0 5 i~ ·co cc..;

~'-
0.5 m cover

Figure 3 Universal design curve Figure 4 Design example 1: simplified method

en
en
=
temperature
100
_ _- - - - 1 . 1 temperature
deg C
100
deg C
80
80

60 60

40
40
20

.
0.02 0.04
I
0.06
I !
0.08
I

0.1 m 0.2 0.4 0.6 m


movement
movement
Figure S Design example 1 Results of level 1 UPBUCK (elastic) Figure 6 Design example Results of level 2 UPBUCK (elastic-plastic)
stress

500
N/mm2
Vo" "",,se .
400 i--;:;-;;-;;:-;;;w<ri'~Y;:~i'";:.::;;:s:..:eq~UI~V~al!!e~nt
0.96 SMYS for X60

300 -1------1'-....

200

100

0
.
8 10 12 14 16 mm
wall thickness
-100

-200

10ngztUdinai
-300

Figure 7 Example of effect of strain-based design

hOOp strlSS

longitudinal suess

a stress patn during loading beyond Yield


assumes; no banding. no rlsiduaJ tensIon. temoeratura increase accliad !'irst. pc-essura increase second

comoresslve
lonQltudinal ~m8S1

b incremental strell Changes' i~cad by bendlnq from Inltiat statl rlcrlsemed :,y :xunt S

Figure a Implications of strain-based design for stress and yield in pipeline

560

Anda mungkin juga menyukai