Vous êtes sur la page 1sur 10

W. V.

BREWER
University of Tulsa, Tulsa, O k l a .

Influence of Lay large Motions i n a Deepwater Pipeline Lail Unler Tension


Tensioning is a prime requirement for laying deepwater pipelines. Of the several new or improved pipe laying methods which have been proposed, or even utilized, all require tensioning of the pipeline to minimize stress in the critical area. In these methods the use of a stinger is optional. Dixon and Rutledge produced charts by which the minimum tension and angle of inclination could be determined for laying pipelines in intermediate and deep waters. The pipeline is freely suspended from the sea floor up to the stinger, or to the lay barge if no stinger is used. It takes a shape over its unsupported length which differs from a natural catenary at its ends due to the pipeline stiffness. In the present work the authors have employed the same mathematical technique to study the sensitivity of a tensioned pipeline to lay barge motions, i.e., surge, heave, and pitch. Two general trends emerge from the results which are presented both in the form of graphs for several sizes of the pipeline and for water depths up to 1000 ft, and in the form of dimensioidess charts. Surge, and to a lesser degree heave, is influential in shallow or intermediate water depths. Pitch becomes a critical parameter in deep waters where, the tension is large. The influence of a sloping rather than horizontal sea floor is also studied in this ivork. It is shown that for a practical range of slopes the prior analysis gives a satisfactory description of the configuration. However, additional curves 'which have been provided are required to make application of the previous results.

0. A. DIXON
Pan American Petroleum Corporation Research Center, Tuisa, O k f a .

Introduction
OFFKHOUK production moves to ever increasing water depths, there arises a growing demand for means of laying deeper and larger pipelines. Many new techniques have been proposed for meeting this demand; some have become practice while others are still in the development stages. Most of these methods have in common one important feature: the use of tensioning in order to increase the unsupported length of pipeline and shorten (or eliminate) the conventional stinger. As operations move to deeper water it seems reasonable to expect the unsupported segment of the pipe to become much greater. Recently, the Lavan Pipeline, "one of the world's deepest, longest and largest underwater pipelines" [ l ] , 1 was laid in 290 ft of water using an "abbreviated stinger" and "constant tensioning," Fig. 1(a). The same organization which laid this pipeline is currently building the world's largest pipe laying barge, which will be equipped with a means for applying tension and will be capable of laying 48-in, pipe [2]. Another contractor, known for laying pipe from reels, found it advantageous to develop a

1 Numbers in brackets designate References at end of paper. Contributed by the Petroleum Division and presented at the ASME Co-sponsored Offshore Technology Conference, Houston,

Texas, May

18-21, 1909,

of T H E AMERICAN SOCIETY OP MECHANICAL

ENGINEERS. 22, 1970.

Manuscript received at ASME Headquarters, January

"torque converter" by which a constant amount of tension could be maintained [3|. It is important to point out that "constant tensioning" helps to reduce the influence of lay barge motions, which is the subject of this paper. Still others have proposed total elimination of stingers. One such suggested technique was to lay the pipeline from the vertical derrick of a drilling vessel, similar to running drill pipe [4]. The pipeline would be run vertically downward through the center of the vessel while maintained under tension, making a large 90 deg bend to lay on the sea door, Fig. 1(c). A slightly different scheme which was patented by Postlewait and Ludwig showed the pipeline made-up in an angular position and then laid under tension, still without a stinger, Fig. 1(6) [5]. The benefit of the sloping derrick would be to reduce the bend at the lay barge end of the pipeline. While technology has grown rapidly, so has understanding of the meclianics of tensioned pipelines. At first, Postlewait and Ludwig suggested that an unsupported, tensioned pipeline would assume the shape of a natural catenary; thus they neglected bending stiffness of the pipeline. Such an assumption is valid in deep waters except at the ends. Subsequent theoretical works examined the shape of the tensioned pipeline including the influence of pipe stiffness. This includes the studies by Garcia, Wilhoit, and Merwin, using a finite element approach [6], and Dixon and Rutledge, using a stiffened catenary technique [7]. At the same time .Stewart and Frazer [8] were conducting experimental, strain gage measurements of pipelines laid with and
AUGUST 1970/ 595

Journal o! Engineering for Industry

Copyright 1970 by ASME

FIG. IA PARTIAL STINGER

FIG.

IB-INCLINED DERRICK

FIG. ID-VERTICAL LAYING WITH "OVER STINGER" Fig. 1 Methods for laying deep water pipelines

without tensioning in water depths up to 295 ft. It was interesting to note their observation that "in the regions of highest betiding stresses, dynamic effects accounted for as much as 1 0 percent of the maximum stress, despite nearly smooth water during laying." Certainly, the unsupported and tensioned pipeline is quite vulnerable to the influences of hydrodyiiamic forces; Stewart and Frazer have emphasized the apparent severity of sea influences on pipelines. An understanding of this entire technology will not be complete until an understanding of its interaction with the hydrodyiiamic medium has been mastered. These influences affect the pipeline in two ways: the first is by direct action of drag and in-

ertia! forces on the pipeline, and the second is through wave or current induced lay barge motions. l i t t l e information is available concerning the interaction of pipeline with wave forces due to the complex nature of the problem. Stewart and Frazer were not able to arrive at any correlation between bending stresses and wave action, despite simultaneous recording of lay barge motions. Theoretical work by Garcia, Wilhoit, and Merwin [9], examined influences of current on pipeline, but this is still an initial endeavor. In order to add further information to this diverse and complex area of study, this paper presents the results of a quasi-static study of pipeline response to lay barge motions. The results may be used to determine the motions for which the pipeline will be stressed to yielding and, subsequently, will be damaged. The study is purely theoretical and utilizes a "stiffened catenary" description of the unsupported pipeline. The original mathematical development was by Dixon and Hut ledge [7], who used an asymptotic expansion technique presented by Plunkett [10] to include the influence of pipeline stiffness. Lay barge motions to be considered quantitatively are those in the vertical plane of the pipeline: surge, heave, and pitch. Maximum bending stresses in the pipeline are most sensitive to these three motions. The remaining three motions, sway, yaw, and roll, are less significant and will be treated accordingly. However, in very deep water, the effect of roll becomes of the same order of magnitude as pilch; this point will be discussed at the appropriate place in the analysis. Torsion of the pipeline is ignored; this could be justified through the observations of Stewart and Frazer. A brief description of the influences of a sloping sea floor has been included.

Definition of Applicability of the Model


The lengthy segment of unsupported pipe takes a shape over most of its length which approximates a natural catenary. Near either end the shape diverts from that of a natural catenary due to bending stiffness and boundary conditions which are not com-

Nomenclature= outer radius of pipe cross section (ft) = water depth at the point of tangency with sea floor measured from the upper end of the unsupported segment (point P, in Figs, l a n d 2) (ft) = water depth under the lay barge (beneath point P, or beneath stinger) in the ease of sloping sea floor (ft) E = elastic modulus (lb/in. 2 ) I = moment of inertia (in. 4 ) W = buoyant weight per unit length (lb/ft) = slope of sea floorpositive if pipe is laid from shallow to deep waters A s = surgehorizontal finite incremental displacement of the upper end of unsupported pipeline (ft), positive convention shown in Fig. 2 heavevertical finite incremental displacement (ft) A(r s , Ao"n = change in maximum static bending stress, a, due to surge and heave, respectively tb = maximum static bending strain, found at point P'/ (in/in) bending strain at point Pt 4,
z

a h 'j o

per end, point Pt (Fig. 2) (ft) Parameters, which are defined in Dixon and Rutledge paper [7] and are meaningless in this analysis, but enter as computational parameter

Independent nondimensianal variables

7/

pitchfinite incremental rotation of upper end of pipeline, point P from position of zero moment

nondimensional parameter representing depth

T.s. 7//

rotations of upper end required to relieve bending stresses at point P due to surge and heave motions, respectively <r D = maximum static bending stress (ksi), found at point P / cr = stress at top due to pitch, surge, and heave, respectively (ksi) a!t = yielding stress of pipe (ksi) X = horizontal distance between point of tangency and up-

S = *
Aeb
e tb

[>V
\c

stiffness factor

= fractional change in maximum static bending stress or strain due to lay barge motion

Dependent nondimensional variables

DT = c As

" = Represents depth under lay barge

D = surge to water depth ratio <]l = heave to water depth ratio D

596 / A U G U S T

1970

Transactions of the ASME

palible with those <>i the natuial ( atenai\ The shape of the unsiippoited pipeline, wheie influence of stillness 1- included, is ailed a "stiffened cateuan In the disc ussion that follows, the unsuppoited pipeline segment issumes the shape ot astillened calenan with a single uuderbend, 1'ig 2 The uppet end / ' , is maintained at the lequned angle for i / e m statu ((aim seal bending moment The statu depth is measuied tiom the sea ffooi to the uppei end ol the unsupported pipeline The uppei end leceives the motions suige, heave, and pitch as small (mite displacements horn the statu position The pipeline motion is assumed to be quasi-statn, thus lgnonng its chnamic iespouse This simplified configuration conesponds to he l a u n g method pioposed In Post lew ail e and Ludwig [~i] (Fig. 1(6)) It is umleistood that when constant tensioning, stingers, n gimbaling devices aie used, the uppei end ot the unsupported pipe does not icceive the motions of the la\ baige hull When, toi example, i paitial stmgei is used (Fig !()) the upper most unsuppoited point is just below the last beaiing point on the stmgei it will expenence sin illei motions than those ot the lay baige With the use ot a paitial stmgei, the slope ot the pipe at the uppermost support will not correspond to the slope tor zero bending moment:. Therefore, static bending stresses will exist in the pipeline at this point. If the static moment: is small, then the results in this paper for pitch may be used to determine its value as well as that due to motion. If, however, the static moment at the last bearing point is large (on the order of the maximum static bending stress), the system is beyond the range of validity of the present results; but it could be examined by repenting the analysis using a modified boundary condition to include the large non-zero static moment.

stresses due to heave and surge are not additive because of the phasing difference between these motions. Making reasonable assumptions regarding the phasing and linearity of the contributing terms, the maximum stress variation can be approximated by the following expression:

Ac |ma: = l(Ao-s)2+ (MufV'-'

(l)

where it. is assumed that the surge and heave perturbations are Acrs. , Acr,, small i . e . , ~ 1 and 1 The effect of the motions on the point P, at the top of the unsupported pipe is more involved. The largest and most obvious influence is pilch. Pitch rotates the pipe an amount yP away from the position for zero moment. The bending stress which results is a linear function of y,, for small motions and increases as the depth increases. This is the tensioned beam (dog-leg) effect produced by the increased weight of unsupported pipe. Heave or surge motions will produce similar bending stresses. These motions change the pipeline tension, and therefore its entire configuration is altered. The new configuration has associated with it a new slope required for zero static bending stress at the upper end. Since the pipeline is still at its original slope, it has experienced an effective rotation away from the zero static position. The effective rotation influences the pipeline in the same way as pitch. If the upper end of the pipe were gimballed, then the pitch motion of the lay barge would not be applied to the pipeline, and the end would be free to rotate when subjected to heave or surge. Therefore, bending stresses at the top would not be encountered and only axial stress would exist. Assuming pitch and surge are in phase with each other and both are about 90 deg out of phase with respect, to heave, then the maximum dynamic stress at the top may be approximated by

Points of Sensitivity
The effect of the motions on the pipeline can become critical at two points in the unsupported segment: 1 First is the point of maximum flexure of the stiffened catenary, near the sea floor Pf, where maximum sialic bending stress a,, occurs. 2 Second is the upper support point P, at the top of the unsupported segment. To simplify the analysis the three motions have been isolated from each other and their influences at Pf and P, were studied separately. Consider first the effect of the motions on the point P; near the sea floor. Pitch has a negligible influence. Surge and heave have significant influences that decrease as depth increases. The

= [(o> + asy- + (<rHy}l/*

where each of these individual stresses must be small (must be less than ab). This must be added to the axial stress, which is self-evident. However, for the range of examples in this study, the axial stress at the top is still small and may be ignored compared with the influences of lay barge motions. Sway, yaw, and roll motions, not in the plane of the static pipeline configuration, are beyond the scope of the direct, quantitative study. It is the presumption of the authors that sway, yaw, and roll will be rather less important in their influence on the point Pf near the sea floor as they appear to influence to a lesser extent those areas already subjected to high static (calm sea) stresses. In general these motions would apply a twisting as well as a bending moment, but. in most cases twist should be small in relation to the length over which it is distributed. At greater depths where the angle setting for zero static moment at the upper point /', is close to the vertical, it is expected that the bending influence of roll is nearly the same as for pitch. Since magnitudes of vessel roll motions are usually greater than pitch for common length to beam ratios, roll would become the more significant factor in deep water operations. The quantitative results in this paper for bending at Pt due to pitch are applicable to roll in deep waters.

Influence of Sloping Sea Floor


Results are stated for a horizontal sea floor. Sloping sea floors were also investigated. For a practical range of small, slowly changing bottom slopes (where 5 < 4 deg), the results of Dixon and Rutledge [7] and those to be presented are an adequate description of the pipeline configurations. Slopes of more than 4 deg are infrequent in most, current areas of operational interest. Depth "D," used to present the results for a horizontal sea floor, is that depth at the point of pipeline tangeney with the sea floor. For practical application when (5 ^ 0, this depth is not easily determined; therefore, a means is provided (Fig. 3) to convert depth reading Dr under the lay barge (or under the upper end of the unAUGUST 1970/

Fig. 2

Convention for positive motions and positive slope of sea floor

Journal of Engineering for Industry

597

solpe is negative (laying from deep to shallow water), the maximum static bending stress will be less than for a horizontal slope at the same depth D; for all positive slopes (shallow to deep), the maximum static bending stress will be close to the value for a horizontal sea floor when the depth I) is taken to be the maximum depth experienced by the unsupported length of pipeline (not the depth under the barge or at the point of tangency with the sea floor). Fig. 4 shows distribution of bending moments for the three cases discussed above.

Examples
Explicit expressions for quantities of interest are difficult to obtain. Nondimensional curves, Fig. 3 and Figs. 14-18, were selected as the most convenient way to represent results in a compact general form applicable to any problem within the scope of study. These were prepared with the aid of a digital computer. The Appendix contains the nondimensional curves together with the equations used to obtain them. While nondimensional curves are more versatile, significant trends are obscured by the generality. To illustrate trends of interest, dimensional curves have been prepared for two of the pipeline sizes used by Dixon and llutledge [7|. For each of the two pipe sizes, examples are worked for both air-filled and water-filled pipe. It is assumed in these examples that the pipe is being laid from an inclined derrick without a stinger.
Fig, 3 Depth adjustment for sloping sea floor

MOMENT USING NATURAL CATENARY REPRESENTATION Pf

MAXIMUM MOMENT

PIPE LINE

Outside diameter Wall thickness Corrosion coat Concrete wrapper Submerged unit weight air-filled water-filled Yielding stress (at/) Moment of inertia

S-in. Schedule 100 8.625 in. 0.594 in. 3 lb/ft 1 in. 44.9 lb/ft 63.7 lb/ft 35 ksi 121.49 in.4

30-in. X-52 30 in. 0.625 in. (negligible) 3 in. 183 lb/ft 471 lb/ft 52 ksi 6224.01 in

The start ing point for each example is the choice of a maximum static (calm sea) bending stress ab. For the 8-in. pipe, two bending stress levels are chosen as examples and for the 30-in. pipe one case is considered. Below are listed the levels of ub: MAXIMUM MOMENT FROM N A T U R A L CATENARY REPRESENTATION HORIZONTAL SEA FLOOR MAXIMUM MOMENT FROM STIFFENED CATENARY REPRESENTATION 8 in. Schedule 100: ab 0.75 ffIt = 26.25 ksi 8 in. Schedule 100: <jb 0.90 a,, = 31.50 ksi 30 in. A-52: ab = 0.90 a,, = 46.80 ksi In the discussion that follows, significant influences will be illustrated with results of the above examples. Conditions imposing the most severe problems will be pointed out. Stress near the sea floor at P f is influenced largely by surge and heave motions. Fig. 5 shows the relationship between the increase in static bending stress versus different values of negative surge (single amplitude) for air filled, 30-in. A"-52 pipe. 2 Negative surge (as defined in Fig. 2) produces an increase in maximum bending stress while positive surge causes a decrease. As an example, point A in Fig. 5 shows that in 350 ft of water, 2.5-ft single amplitude surge would increase the bending stress by about 5.2 ksi such that the total bending stress at Pf is equal to the yielding stress (ab + Au,, = 46.8 + 5.2 = 52 ksi). In shallower water, 2.5 ft of surge would give rise to even greater stresses, while in deeper water the influence of surge is less significant. For small motions in deep water a careful examination of Fig. 5 shows that a nearly-linear relationship does exist between surge A,s and stress change ACT,, for any given depth in the range of surge noted. Taking advantage of the presumed linearity, the family of curves in Fig. 5 can lie replaced by one curve. The
2 It should be noted that the method of analysis used here was developed for deepwater applications and becomes increasingly unreliable for shallow water; hence, the curves have been terminated.

NEGATIVE
Fig. 4

SLOPE

Influence of sloping floor on maximum bending

supported segment) to depth D used in this analysis. For a negative sloping sea floor, the abscissa in Fig. 3 is similarly negative. Two general statements can be made about the sloping sea floor problem for all slowly changing slopes small or large: if the 598 / A U G U S T 19 7 0

Transactions of the A S i E

100

|
t

0 8"-crb=3l.5KSI, Acr b = 3 . 5 K S I ^ 100 200 300

i3"-<rb =26.25 KSI, Acr h =8.75KSI \

200-

.|C

^~tn" ^ \"
A l i i UK

LU UJ UX rQ_ UJ Q (T UJ

SURGESINGLE AMPLIT JDE 300t ^ S (FT.)

WATER FILLED 30"Acr b =5.2r :si. 4 0 0 AIR FILLEDzS 500 600 700 800

*%^;
30"-Acr b = 5.2 K S I / WATER F LLED !

400500600700 800 900


1000. 0

-2 -4

-y
A
3

i/
TOTAL BENDING / S T R E S S EQUALS YIELDING STRESS

-10 -12 2 CHANGE

yl
r

5.2 6 8 10 BENDING STRESS-AC^ (KSI)

nnn -4 -6 -10 HEAVE-SINGLE AMPLITUDE AH(FT)


Fig. 7 Comparison of heave influence on bending near sea floor

Fig. 5 Influence of surge on bending near sea floor air-filled, 30-in., X-52 pipe

PITCH - S I N G L I 100 200 p


UJ UJ

1
I AMPLITUDE- fp 3 4 5" 6

I*

no

\
300

i. x

400

1
\

500
600
700 800 900 I00Q 0

5
<
5

\
\

\
10 20 30 40 TOTAL STRESS AT TOP DUE TO AXIAL LOAD AND PITCH-(KSI) 50

-2

-4 -6 -8 -10 SURGE-SINGLE AMPLITUDE As(FT)

Fig. 8

Influence of pitch on bending at top air-filled, 30-in., X-52 pipe

Fig. 6

Comparison of surge influence on bending near sea floor

curve for air-filled, 30-in. A"-52 pipe in Fig. 6 shows the amount of (negative) surge at each water depth which is necessary to increase the maximum static bending stress to the yielding stress (i.e., Ao-j = 5.2 ksi, or <rb + Acr,, = 52 ksi). The example noted above for 2.5 ft of surge in 350 ft of water is noted by point B in Fig. (3. Also shown in Fig. 6 are the surge versus depth conditions that produce yielding stress for the several other examples. Note that

air-filled pipe is far more vulnerable in the case of 30-in. pipe. In the case of 8-in. pipe, air or water-filled makes little difference. However, the S-in. pipe with the lower static bending stress, ab = 26.25 ksi, is far less sensitive to surge than with <rh = 31.5 ksi. Fig. 7 shows a similar comparison for heave. As with surge, static bending stress is increased by negative (downward) heave; furthermore, the influence of heave decreases with increasing water depth. Comparison of Figs. 6 and 7 shows that t he siresses near the sea floor are less sensitive to heave than to surge. Stress near the top of the unsupported pipeline segment, at P AUGUST 1 9 70 / 599

Journal of Engineering lor Industry

is influenced most by pitch motions. Fig. 8 shows the total stress versus pitch for air-filled 30-in. pipe; bending stress due to pitch has been superimposed on axial stress to show the relative influence of pitch. It is apparent that the stress due to pitch increases moderately with water depth, as opposed to the influence of surge and heave at the bottom which decrease with increasing water depth. The bending stress is a linear function of pitch at each water depth for small rotations, say yP ^ 10 deg. This is easily concluded from the formulation in the Appendix. Therefore, it is in order to represent the pitch influence by a single curve for each pipe size as is done in Fig. 9. Fig. 9 gives pitch versus depth

100

200
rUJ UJ

-2 -~~"""

300 400 500 600

77/ /^~~

x a.
tLU Q

I // loy i '
12

cr
UJ

100 200 P
UJ UJ

8"- cr b = 0.75cryp=26.25, 0>=CTyp = 35KSI

-AIR FILLED

HEAVE- JDE SINGLE'AMPLIT 700 A H (FT)

V 1
D /

800 900 1000. 0


F i g . 11

300 400

cr b =0.9cr /P =3|.5 i crfc=cryp=35KSI

/ /

uCL UJ Q CC UJ I-

10 20 30 40 50 BENDING STRESS AT TOP ( K S I ) - a b t .

Influence of h e a v e on b e n d i n g a l t o p a i r - f i l l e d , 3 0 - i n . , X - 5 2 p i p e

500 600 700 800 900 1000, 0


Fig. 9

WATER F LLED

I /

<

7
\ \\

-\

1 1

4 6 8 PITCH (DEGREES)-Xp

\A f \

30" cr t =cryp = 52 KSI "WATER FILLED^ 30" 0-t=cr yP = 52KSI AIR FILLED

'

10

C o m p a r i s o n of pitch influence on b e n d i n g at t o p

0 100 200
rlll UJ

300

^-""'
400

u.
X H-

4
500 600

a.
Q

6
8

10 20 30 40 BENDING STRESS AT T0P(KSI)-crbt FROM 10 FT SURGE


Fig. 12 Comparison of surge influence on bending at top

50

rr

A/
/

MI r-

< 700

10 3URGE800 -SINGL E AMPLI TUDE L\ S ( F T )


c

900 1000. 0
F i g . 10

10 20 30 40 50 BENDING STRESS AT T0P(KSI)-b t

Influence of surge on b e n d i n g a ! l o p a i r - f l l l e d , 3 0 - i n . , X - 5 2 pipe

conditions that produce yielding stress for the several examples. Axial stress has been excluded since it is small compared to bending stress. Fig. 9 shows that air-filled pipe is less sensitive to pitch than is water-filled pipe. This is because the tension required at the top to support the heavier, water-filled pipe is far greater than for the air-filled pipe. Bending stresses at the top due to surge and heave are shown in Figs. 10 and 11 for the air-filled, 30-in. X-52 pipe. Surge and heave influences are strongest, in shallow waters where pitch influence is small and decrease as depth increases. They are un-

600 / A U G U 5 T 19 7 0

Transactions oS the AS ME

uiipnitinl at mteimediate depths e\<ept in u e e - ot (oml)uted d i e d - ot all thiee pitch, heave, anil -uige An example i- di-i listed below Figs \2 and 1,! aie given to compact1 the lelative influence of he ive and -uige on bending at the top foi the seveial -ample pipe'meKadi u u v e -how- the bending -tie-s art veisti- watei depth due lo 10 ft ot -mge, oi heave The lapid d e n e a s " of this mflueme t u t h iiuica-iiig witei deplh i- \ f i \ obviouVl-o, as ni the i a-e ot pitdi, the heaviei pipe -m h a- the watei-filled, ,!0-iu |ii[)eluie l- mm li moie -en-itive at the lop to la\ baige motion1>\ aummg the Imeai lelationship ot -tie to singe, oi heave, i erchwatpi depth, it is polble to deteimme the bending - t i e due to am amount ol-in ge, oi heave, lioin Fig- 12 ot I ! \ n eximple l- given a little tuilhei m tin- papei The le-ult- loi pitdi lepie-enl the bending -ties- caii-ed In a hange in angle at the top ot the pipe Vn\ augulai deviation, eiMidle ol -Dinip, mil impo-p (he ^dnw -lie Foi this iia-on, the le-ult- ma\ be used to -tud\ the of let t of anothei [iipe-lme-Ln mg (ondgmalion, Fig be), pio[iosed loi clpppwatei ipeiationLa\mg pipe lioiii a veitnal pn-ition, lontend thp piopoupnls, is thp e t-ie-t and best method ol leduung stiees leultmg Imm la\mg ])ipe m dee]i watei The I'ipe make- ouh me long ladiu-, '10 deg bend instead of the iS'-cuive mipo-ed b\ 1 iMng the pipe horn the hoi i/onl d dec k ot a 1 i\ baige ' [1] With an adequale hoii/ontal (ompoiipnt ol pipeline len-ion to lUPVPiit lailuie neai the -ea llooi, the pipeline would \ leld at the uitace when' Hie veiricul -lope di/fei- con-itk'iabh fiom that lequued lot /eio -lain moment Foi all example- u-ed m this ~'uch, thp inimmum angle foi / e m bending moment M the top l^teatei than 15 deg [7] loi 100 l-tl watei oi le It I- appaient mm Fig (l that \ieldmg OKIUS in e \ e n ca-e, at auglp- le than ilns Thpiefoie, the piopo-ed method l- inadequate II (ould be modified to unhide a paitial o\ei--tmgei as shown m Fig 1 (/) Vn ovPi--tingei would (ontiol the bend neai the -uiface uid peimit the pipe to be aentitled \ e i t n all\ a- pioposed she neai thp tup lau-pd In a combination ot motion- could oe -igmlnant at mteiinediate watei depths wheie each motion individiialh might not le-uli in Meld ofthe pipe A.s an illu-lia'lon ( on-idei tlieS-m Sdiedule 100 wilh a maximum static bend-

ing stress ab equal to 20.25 k.si (0.75 at)). The pipeline is air-filled and being laid in 200 ft of water. Assume that the pipe is beinglaid without a stinger by the method shown in Fig. 1(6) and that lay barge morions are: surge, 4-ft; heave 4-ft; and pitch, 3 deg. The surge-heave influence at the sea floor is comparatively small. Fig. fl, point C, shows that 6.3 ft of surge (only) would produce a change in stress of 8.75 ksi at the point of maximum flexure. Assuming linearity of stress to surge, the maximum bending stress due to 4-ft of surge would be / 4 fl \ Ao-WsllrKc = I T-r~r I X (8.75-ksi) = 5.0 ksi \().o t y Using Fig. 7 in a similar fashion shows that 4-ft of heave might produce 2.4 ksi. Since pitch does not influence bending near the sea floor, the maximum dynamic bending stress is about Ao-i,/,,,,, [(5.0) 2 + (2.4) 2 ] ! 6.1 k.si,

and the maximum total bending .stress at the bollom is approximately 32-ksi (20.25 + G. 1 = 32.35). At the top end of the pipeline, point O on Fig. 9 shows that 5.7-deg pilch produces yielding stress in bending, 35-ksi; consequently, 3-deg pitch will give 3 deg
C(/l)itcli =

5.7 elegy

X (35 ksi) = 18.4 ksi

Using Fig. 12, point E, surge will produce a bending stress of oYsurse = ( ) X (47 ksi) = 18.8 ksi and from Fig. (3, point /'', heave gives 4
Cl/lllMVC

10

X (If) ksi) = 0.4 ksi

In most instances the phasing of surge and pitch will be such that their influences are additive; these two alone will clearly cause yielding of the pipe at its lay barge end, a, = 37 ksi. The maximum dynamic stress is
fff/n

[(18,4 + 18.8) 2 + (6.4) 2 ]''- = 37.7 ksi

Conclusions
1 In deep water, only pitch and roll have a significant effect on the unsupported pipeline. Pitch is more critical with heavier pipelines due to the greater tension at the lay barge. Therefore, water-filled pipelines are more sensitive to pitch. 2 At intermediate depths, surge and to a lesser extend heave are the significant motions. Their influence is greatest at the point of maximum static-bending near the sea floor. Increased tension and the use of constant tensioning devices and partial stingers should reduce this problem. 3 Combined effects of surge, heave, and pitch at the top of the unsupported pipeline could be significant at intermediate depths. Increased tension would not improve this problem; however, constant tensioning and partial stingers should help. 4 Slowly varying bottom slopes of 5 S! 4 deg do not change the results appreciably; however, the input parameters must be adjusted with supplementary materials provided to apply the results of this study. 5 This paper can be used to (a) determine maximum barge motions compatible with the integrity of the pipe; (b) ascertain desirability and usefulness of now barge designs with lesser inherent: motions.

fLU UJ U. X hQUJ Q

tr
UJ

l000

10 20 30 40" BENDING STRESS AT T0P(KSI)-crbl FROM 10 FT. HEAVE


Comparison of heave influence on bending at fop

50

Acknowledgment
The authors wish to thank P a n American Petroleum CorporaAUGUST ! 970 / 601

Fig. 13

Journal of Engineering for Industry

ticm for tlie authorization of releasing the results of this study. They are grateful to Arthur Lubinski, Harold Quellhorst, and Dr. A. E. Knapp of Pan American for their valuable advice and constructive criticism, and to J. A. DeVoe for his assistance in preparing the graphical results.

a
r'/i

>b)

The angle of the pipeline at its upper end which is necessary to eliminate bending moment is
ah

References
1 O'Donnell, J. I'., "Offshore Line Laid Under Tension," The Oil and Gas Journal, Vol. 06, No. 20, June 24, 1908, p. 82. 2 Offshore Newsletter, Vol. 1, No. 7, October 23, 1908. 3 O'Donnell, J. P., "Offshore Pipelining: A Special Report," Tlie Oil and Gas Journal, p. 07. 4 O'Donnell, J. P., "Suhsea Pipeline Challenges are Depth, Cost, Distance," The Oil and Gas Journal, July 10, 1907, p. 125. 5 Postlewaite, W. K., and Ludwig, M., "Method for Laying Submarine Pipe Lines," Patent No. 3,200,250, United Slates Patent Office, August 16, 1960. 0 Garcia, D. J., Wilhoit, J. (.'., Jr., and Merwin, J. E., "Bending Moments Induced in Laying Offshore Pipeline Under Tension," ASM E Paper No. 07-PET-8. 7 Dixon, D. A., and Rutledge, D. It., "Stiffened Catenary Calculations In Pipeline Laving Problems," JOURNAL OF ENOIXHKRINO FOR
INDUSTRY, TRANS. ASME, Series B, Vol. 00, No. 1, Feb. 190S, p.

0(1) , = t a u - ' O ) +

oi

The water depth D (Fig. 2), measured vertically from the point of tangency to l' is expressed by

-- = ur- + i)'-'! - (A2 + .V)'-A


+ a1 1 _ h*
IS)

An expression for the horizontal distance X between the point of tangency and / ' , was obtained using the same method as for deriving D/L in reference [71.
A"

153. 8 Stewart, T, L., and l-'razer, .1. P., " Experiment a Measurement of Stresses While Laying Pipe Offshore," ASME Paper No. 00-PET-24. 9 Garcia, D. J., Wilhoit, J. C., Jr., and Merwin, J. E., "Current Induced Bending Moments in Laying Offshore Pipeline," AS.ME Paper No. 08-PET-O. 10 Plunkett, It., "Static Bending Stresses in Catenaries and
Drill Strings," JOURNAL or ENOINEERINU FOR INDUSTRY, TRANS.

= /(

Slr

a2h -I +

(h + \f-

(ft + \Y * '

ay - :

(9)

Sloping Sea FloorFig. 3. The previous equations are for a horizontal door. The sloping sea floor alters only one, MS): the equat ion for z() is replaced by 2o = - , T72 - Sh
(10)

ASME, Series B, Vol 89, No. 1, I'eb. 1907, pp. 31 30.

APPENDIX
Nondimensional Curves and Pertinent Formulas. The basic mathematical technique used in this analysis is the asymptotic expansion as originally developed by Plunkett (Hl| and then specialized for unsupported pipelines by Dixon and Rutledge [7|. The same assumptions and limitations governing those works exist for the present results. Additional assumptions will appear in the development as necessary to achieve the aims of this paper. This Appendix contains a summary of those expressions which were used to prepare the nondimensional curves, Fig.'>and Figs. 14-18. A complete derivation of these formulas can be obtained from the authors upon request. The calculations have been greatly simplified by using an empirical result from the Dixon and Hut ledge paper (reference [7), p. 155): the required horizontal force II applied to the pipeline is very little affected by bending stiffness. Therefore, this force can be computed with sufficient accuracy by the natural catenary method,
II

The expression used in preparing Fig. 3 is I),. = I) + StSh'^Ul')1 + '2D)l'"-\ + S' = + o>
ill i

It is obtained strictly on the basis of geometry, see Fig. 2, where equations (8) and (!)) were substituted for I) and A. PitchFig. 14. The expressions in reference [7] were developed for a zero bending moment at the top. The angle of the pipeline necessary to satisfy this condition is given by (7). The expression describing the shape of the unsupported pipeline was rederived, changing the boundary condition at the upper end to conform to a prescribed slope

10.0 5.0

X
.10

cW
6

()
1.00 (4) 0.50-

This result is used extensively throughout to avoid the simultaneous solution of two transcendental equations for the location and magnitude of maximum static bending stress eh. This empirical relationship is valid for intermediale and deep waters, say 100-ft or greater for the 8-in. pipeline and 200-ft or greater for the 30-in. pipeline. The following expressions from reference [7] were used to derive the formulas from which the nondimensional curves have been prepared. At the point of tangency with the sea floor the pipeline makes an angle 0 with respect to the vertical given by Q{z) = t a n - 1 where za is approximated by 602 / A U G U S T 19 7 0
ah. (o)

B-

DD

.ci4
0.10

.02

/ / .01 /

0.05

/=v
0.01,

Iffb)

3 .00' r 00l|

0.2

0.3

0.4

0.5

0.6

J_fbt Yp b
Fig. 14

_[m./ nr:r\ llN/DE6-/

Nondimensional curve for strain at top due to pitch

Transactions of the ASME

SurgeFigs. 15 and 16. Positive surge A s will increase the horizontal component of force and increase the unsupported length of the pipeline. Therefore, the maximum static bending stress 6,, is decreased an amount Ae6. Conversely, negative surge will increase th by an amount Aeb. There must be no heave or pilch so as to isolate the surge influences. Fig. 15 was developed by first applying a small fractional change Aei/eb to the static bending stress; the configuration of the unsupported pipe is altered. Then the surge AA. was equated to the resulting incremented change in X (9) while slope (12) and depth I) (<S) were maintained constant. The steps are as follows, where ys replaces -/,, in (12): (1) L, is found from the implicit expression D

! _ , ' '

"+s\6
+

[TTsF

(14)

(2)

/ . _ > and yN are solved simultaneously from

L,' " = Dc + 11 + Sf-3],/-

s
Fig. 15 surge Nondimensionai curve for change in strain at bottom due to

"Mi [r+W

' Sc;l ' (7s)

(15)

and ys = t a n - ' [(L, i r ' / ! ] - t a n - ' [(L D~'A] ' Se-'_' L25/=

s
where (3) Ae6 e = 1 + <. A, . . Finally, -r is found by substitution into

(16)

0- ebD

|-(^)sfc-M(L,-l)V.]-(f)sfc-M(L,-l)V.,
_Sc=

( U _ - l) 1 '' 5

s
L,1

(L -_!)'
_

~b7"

D ~

10.0 5.0 Aeb eb \ O.bV

Q2

^K^
N \ \ __^,. / \

Fig. 16 Nondimensionai curve for change in slope at top for zero bending due to surge

N \\ \ \ \
\

1.00 0 ( 1 ) = 9 ( 1 )|,_ + 7p
(12)

0.50

'A
/

\\

-f
~ " /

where y,, is the angular deviation from the slope required for zero moment. If the pipe is clamped at the lay barge, yP equals the pitch of the barge. The moment in the pipeline at the top is it linear function of y,,. Using this approach, the bending strain at the top e,, as a fraction of the maximum static bending strain at (he bottom e, is given by

fi'^
0.10 0.05

, j j
/ / ^ - ^ -

hy_

JD+

t-H,m^ I 0.01
0.001
0

(13 0.01 0 -.01 -.02 -.03 AH/D


Fig. 17 heave

!
-.05 -.06

Fig. 14 shows a family of nondimensionai curves developed from this equation, where the abscissa is the fraction I J per degree

-.04

Nondimensionai curve for change in strain at bottom dve fo

of pitch y,,.

Journal of Engineering for Industry

AUGUST

1 9 7 0 /

603

Since no slope change is permitted during surge, a moment is applied at the top of the pipe which is equivalent to applying a pitch 7 ; , = ys. The resulting bending strain at the top is found entering Fig. 14, for pitch, with the angle 7, s . HeaveFigs. 17 and 18. The influence of heave is analogous to that of surge. While maintaining the horizontal position and slope of the pipe constant at the top, a small vertical displacement upwards will increase the horizontal component of force and the unsupported length. The steps for calculating heave are similar to those for surge: (1) Ae,, Given a value of ', Li is found from ! 11).

(2) By replacing 7.s. with y in (1(5), L2 and y are solved simultaneously from (16) and (L, l ) l / s = c(L, l)1'-- + S/i-|(L2 - cS/r'KL, "Sr ! (L. - 11 i) 1 '-'] l) 1 -''
1

l)''''=] S(L, IV

u "

+ (7
Fig. 18 Nondimensional curve for change in slope at top for zero bending dueto heave

Sc (L. -

CIS)

(o)

Finally, , shown in Fig. 17, is found by substitution into |L 2 ]'A - (l + Sc=)'A

' t U - l) Dc

(L,

- n'Ai D "Sr'(L . L
D"

I) + & (17)

+ (7.s)

'\
Dr)

(1 + Scs)''

(7

r s<-

(10)

The computational parameter 7,s., shown in Fig. 16, represents the change in the required slope for zero moment at the lop.

The parameter y, Fig. 18, is the change in required slope due to heave. It has an effect equivalent to applying a pit eh y,, = -In-

604 / A U G U S T

1970

Transactions of the AS ME

Vous aimerez peut-être aussi