Model tests and analysis of flexible riser systems

;r a catenary flexie riser system is

stucie:

reat.:r1 to

a tYpical field develccnient in

the cr:n Sea. Sucn systens exnibtt enavIcur wflich la

ocre :c:piex en:

ffìou:

t:

:eternlre toan for a

riI: vertical rcser.

The pacer fIrst preser.ts simole and epprcxnate fcroujae Ievelce: by :atenary theory. Thteoary sclutcn aCcuratelï esorLbe the shape,

cel

sic :oclliaticns cf a riser

_f

its flexural

stffness is low. fornaI asynptctc expansion is :0er. _{precente: and this netoca :s aop.:ea to the}

non-trear :ehavto.ur of a lcr, flexible rIser.

The pacer

:ee.:os tre governine equations for the non-linear

:yrani: aralyss cf a flex:ote rcser systen and cutlines one nethod of solutIon casco On a

fl.ilte

element apprcaoo. . sertes cf nooe tank tests vere :arr:ec cut ..«. a wave f!u.-.e

i

croer t: coserve the ocr-.ioear response cf toe fiex.ble -15er svsten. The

systen 000s:sted cf

a fexì.e ctpe

suspended to a

;stenarv :e:eeî a subsurface cuoy ana a surface lucy. Tres e tests- e-e oestgre:

t:

cover :tfferent waves ing coto cperatir.; so: severe ervir:nrr.eota cc ri: lt c r. s

3 cors tan t

fluid crag coeffIcient

added nass ccefftc:er,:

cuter Oieneter cf toe riser lendne r:gidit/ cf loe riser ct pe

external forces co Iwo v,w orthogonal cireotton

t:

axis cf rIser element

respectively

ncrzoo:al compcnent cf effective riser tens i or.

non-:lnvnslcna2. ootatcn

lenaIr; occen:

nass per cr11 length wr.to. Includes rIser nass, nass of :r.ternai fluid ano

adde: nass tern

arparen: wecgnr. of che riser plus concerts per unit :eng:n (viz actual

MODEL TESTS AND ANALYSIS OF FLEXIBLE RISER SYSTEMS 0.0. Owen and K. Qn

s OUsore inrserr

Kennt Watt Unverety

Esnbigr. U.K TECHNISCHE UNIVERSITEIT Laboratorium voor ScheepshydrOmeChafllCa Archief Meketweg 2, 2628 CD Deift Tel.: 015-786873- Fax: 015-781836 354 Greek Synbcls a nass matrIx dancing matrix stffness oatrlx fcrce catrix

displacement, velocity and acceleration

focal acplltude matrtoes non-dneosional notation

slope of riser curve with respect to toe vertical

censjty cf sea water 1 Ia

Superscr:;:s n 'atrix Ecuattcn assencle: oa:rx

elenent.al matrix Sucs:rc;:s In Matrcx quation

giocaI ccorOinate systen

local ccordrate system

weignt less buoyancy force componen:)

s total length of riser

s length cf rcser

effective riser tension

horIzontal component cf effective rIser

ensicn, T0

1 vertical ccnocnent of effectcve rcser

t e na ion

j

a' relative velcc:ttes in two crtr.ccnsì normal direotons of tne element of toe

iser

v,w deformations of differential elenco: orthogonal to axis of riser

fluid aooeieratons in two orthogonal ncrmal direoton cf toe element of toe riser

components cf riser apparent veigot orthogonal to axis of riser see

cefiniticri for P)

Y Yo non-dimensional notatlob

-sr;i.ial fiels srS Ire

:_:i: :lverticr.ai felSs 1:a:;r ?rccuCtcn

)F?5) ere

3.

..IrcreasLr; ere5t frtm anS aS rS_stry. Tley raye le ccnstSereS

-,--n::ai leveL:eents s:rce :r.er vantages ere

:ant r.er ctier crcSuc: cn sysers as f xe

.s a s-; :crcnent r Ire SC5ifl

v atir systen. re are :esi:l1y t-c

:J;es a:-Lte :-3ers:

re r:;i steel riser;

; :re flexi:le r:ser.

.ax;:L e riser s :rcreasln gly 5e:r adcpteS fl

-ecr years. In 978 :r.e first fìx:ILe r.ser .ias vccessfu1lJ r.trcduceS i.n te rcncva F:eLS in

razL ..n 179, f1sxilIe rsers .ere usec Ic

ti! S3a.n. Since titen lite flex.ble r:eer teen useS t- '/srcus arts cf Ire crlt, ecn as

-e :55130 FielS tn tie ?tLitpines

ref. 2), tre 7:er.3 FIeLS .n :r.scresia, etc, anS tifferent types cf :extlle risers ere .-e:cpeS f:r Stffer5nt FIeLS eoLtcacicr.s (refs. 3,-,5,: . -icre candly several

;cmcanles raie prtDcset flcattrg rrcducticn

system

:cnceo:s usi:-.; flexilje rtsers Fc:- future Ncrtn Sea cl t5 ;ss prcSuctcn. cflextp (ref. .5) :resentec a

tv:t:aL Sevelcpnent .nvoil:r; FlexIcLe risers fcr a

:1t3::rg trcSucdcn fs" ; _{y fc:-} tre Ncrdn Sea. ast

¡car, foster treeler levelopoent (ref. 7) presenteS

Ire;:- flexitie rtser system for a ae-sumers;SLe

'y. It atout cre sane dIme crley

Sr. r.esr:r.; ;ut fcr.arS

zn-a::-f?3 concept cnoloytr.g a

:..CXLDLC r;.ser, enowi s cre caro er:i null platform (ref.

i). Taylor cccrc- lffsrcre eveIcpment rcup raie recer.tiv presenteS a turret ancnor prcoucticn system

TAPS). :c enoulS also te mentioneS dha'. ha.niLtcn

l:-ttrers ra/e a fexltle r:aer Fc:- datar :ryacttCn cn

cre eetsea P;or.eer r;; tn tre Âr;yl f;elS (ref. lIso, ::or:n Sea Sun .111 :c ts rave saiS trat flexitte risers dcult ce used in :r.e lalmoral ?e_i.

Tre ;rcreaslr; :nterest fexttle r;ser systems

Sea oil and ;as 5evelopcents La nardly

uurcrtstr.; In view cf tre attenticn cetr; gIven Io tre tevelcpment cf small arS ceepwater Fii5s. cward Pass

ref. "), in locking st zne potentLal cf us to 0 rew TC!SS wh;cr could e Seielc:ed on Ire < or.tir.entaì :re:f cver cre rext 15 co 20 years, trcioated that tne ra;crity cf :rem were Likely co te nar;tnsls unoer 100 rt.Ll:er 05:-rel reserves. floating production syst?OS

s;r; Ftexzt!e r:aers offer sin;fitart advantages over otre:- :rcductccn systems. crtffitrs ref. 12) 1iscussS the asvanzage cf fexttle r;ser systems tn-i

000;ar:scn wttit r;gLS rzsers. Shcttclt 'ref. il) rcnsiSered the overaLl performance cf prcductiCn

sfslams arS also 3lscussed tite rajor asvantages cf

F.ex;cle rtsers cver rt;5 risers. Two cf tr.e matti

atrantages cf flcxi.5lC rtser a/stems are:

they are aoje dc accommodate mottons;

t) trey have relatively low irstailaticn and operating

0t5t3

This paper axamires catenary rtsers wnicn consist cf cign pressure f!exttle pipes ranging in a caterary curve, corrected st tre uPper end to a facilIty ceck,

pas3irg cver a tray supported sy a autaurface oucy and

runnIng vertIcally Sown to a rtser case cm tre seated fig.

i-,

:n :rnciple, a ocre preetse analysis cf a

riser would reed to tace Lrto account its cer.Sir.;

st:Ffress. ano tris wculî lead tc a so'-' -r cf

ncr-lirear SifferenItal suat:cns. This aoluticn cari :r. practice cnLy te carried out Dy numerical nears us ir.; a Large computer. eaPite trese :cnpLxtt:es. zeno;-. st:ffnesses are ;eneraly relatively small eno s:rcle

anS sprcxcmate fcrmulse can te used, sucn as caterary

theory. The catenary solutions ;tie accurate

descriptIons cf the configuration and tsnsj-cr.s

Fc:-flexttje rIser systems.

Fcllcw:r; tre iscuaston of ostenary acluz:cns s Fcrmal ssynptctic expanstcn, wnlcn is valid to cescrtcs

the Seflection cf en unsupported pipeLine Includi:-.; flex-urat stiffness in Seep water, is presenteS and applied Io a flextole riser. Simtlar cetrioSs rave teen

used in pipeLire laying prcclems )refs. 14,15).

FlexIble rsers exh;tit Synamtc lena-/t-cur wntcn :3 core complex and Stffioult to determine Iran trat

for r;gtd vert:cal r:sers. yodel testa are trerfcre

needed to predict thC performance of a flexiole r:ser. :ri view cf the relative scarcity

of model Sate, a series cf model tests ere carrteS tut cy tre autrcrs,

wnioh -will be supplemented by Future more cetailed experiments. Some cf t:re initial test resulta

are presented ini this paper.

ANALYSIS TF FL:xIBLg SISYA STSTMS

Trie flextole r:ser system consLereS :n tris paper

;s typIcaL of trose proposed Fcr 'use ir cre ortn Sea. Thts nypotnetloaL Fteld is assumed to rave CO nLll:cn barrels reserve enS is to te cevelopec cy tears cf a floating prcduct:on aci!Lty. The water Septn :3 :5Cm.

The flextol rtser system )?;;. ) consista cf oressure fIex:cLe p;pe mangi:-.; cetween tre ceoe cf

tre

rig and a suosurface cuoy. The rtser case la CC away from the facility. The subsurface buoy is sttusted

SCm

above the seabed. The riser system is to ce esi;neo to withstand tre 100 year sLcrn.

rTENARf sCLiJ-r:CN

The flexible riser system ras reLat;'/eiy low

tending stiffness. As a consequence cf Ints, tre

oonfigurat;cn cf tre rtser ta easenttaLly cf a catenery

form. The bending stiffness is likely to have

Lnfluence cri the deflection tier mcst of its langor so trat the deflection arid tenstcns cf lite rtaer oar, ce calculated wtth aufftotent accuracy sy re;lecttn; tre tending stiffness arc using treoatenary ac-,,n. The shape of the flextOle rIser will also te :r.fluencec oy axial FLexltility. cwever, it can oe crown trat tre

error Invclvec in assuming that tre rtser ta inextensible ifrom the point 'cf /ieu cf tetermtn:n; :ts oonftguraion) ta small tri practice. If tre effect cf axial flexibility is t-o be tacen Into ccnsioeratcon. one could neec to resort to numerical methcOx cf

acivin; the :fferentiaL equatcns of ata:tc

equilibrium Fcr the flexible rIser.

The .efleotion of tre catenary is given Dy;

-z

- Acosric

where A

T.0 - horizontal component of effCctiVe riser tension

apoarent weignI cf the riser plus

Applying this to a f1xib1e r:xer gives Fig. 2)

x-x

- Accsn---" o

(2)

equat:on 2) _{tre parweters A, x0 and o depend} cm tre :ser :egtn used, as well as the positicn of tre pcirts and the riser apparent weight. This ts the actual riser i.eignt 7l..s

tre welgt of internal fluid

lass tre cuoyancy forca component. when tre length S Ls

i,titese :rameters a

defIned and trus the snape

of the riser s gj'en by

(-X X

z AEOOSh __2. ccsn (3)

The slope, (x), can be obtained from equation (3)

x-x

-1, .dz -1.

(x) -

tan U') -

_tan

t/stnn -;

dx 0,

ThC effeotve tnston, T(.$), is ven oy

2 2

:(s) o? s-l' T

P 0

and the length., 3, :s gu ven ty

x-x X

s A(sinh _A ° s.nn

(U)

:n equation (5) above, the ocncept of "effective

tensIon' _{:5 used :n order to tage S000unt of :nternal}

pressura effects and components cf force arosing fron

external fluid pressure (viz througn application of Aronimedes Law). An excellent descrlpticn cf tne

concept of

"effeottve terstor' arO its application to

riser OefornatiOns and stresses is given in a 2aper by

Sparks (ref.16).

The physical characteristics used for the results presented are g:van :n Table ',As already discussed,

tre :ength cf tre riser, 3,

:s an important factor. S

will te within t're range Si K S K 32 wnere S _{15 the} linear distance between the deck and the suoaurface ouoy arS _{2 is determined either by the riser contacting the} seabed or by the cinimitli allowable bending radius.

sually:he mininn bending radius is relatively small (see Table 1) Snd hence

it

is the contact with the

eeaoed that normally determInes Th. Fig. 3 snows how the cor.figuratcn cf the riser :s :rfluenoed by the riser length, and Fig. A snows the corresponding values

i'cr riser tension. _{The minimum tensions T1j(-T0) are} also affected by the length, but the maximum tensions,

Trax, are muon less affected. _{This indicates that the} water depth is a ocre important factor in influencing tension at

the rig

tr.arl ta

tre riser

ength. 'ave-induced notion of the rIg would of course affect the confLguration

f

the riser,

fig.

5 3hcws riser ccnfgurationa for different values

_{cf horizontal rig}

;.csitlon, expressed as a percentage of 'dater depth, for a riser of length 200m, _{It should be noted that rIsers} with different lengths

'dill

have _{different natural}

frequencies and in some situations _{ignIfioant dynamic} effects could occur.

ST:FFSNEO CATENAg SOLUTION

More accurate solutIons can be _obtained by

considering the bending _stiffness. An asymptotic expansion, whion is valid to describe the defleotton Pf an 'unaupported pipeline includIng flexural stiffness in deep dater can be applied to the flexible riser.

The equllLbrt'm equations for the element of r:ser sr.o'wn In Fig. 2 are srcwn below.

356 Mocsi - Vsin8 - J Ss -- const (M - T,) cd - -O or '-'s il!._os - or 't s 'i p _p -1 and dz -i-x - oosd

, -

sirU Os ₀₃

From classical ocam tneory 59

-i

_os

Jsing the nor.-d:mensional rotation

- (s/S -h - _{dl? 3} 2 2 - s3 p equation (7) becomes

'ii

a -'

hoosU - yxird - S s'y

1f tre sttffness

is reg-,

2 _{- O and the}

equation U2) then leads to the catenary so.uticn. In the case of a stiffened catenary, 2 t then an _{asymptotic expansion given by yt-unket: (refs.}

!i.15) will give an approximate solution. _{The solution} is Oescribed by its angle with respect to the verticalI C(y) - I(y,a) _{01(y,$)exp[-s(y-y0)q1(y)}

s,(y,a)expE-m(y0,-y)q,(y) (13)

where o

_'/5.

The first term cf the above solution is _{a serica} expansion whon satisfies the cifferential emuation (12)

but not the boundary condItions, wnioh _{are satsfed by} the remaining terms. _{The first term oan be dritten In}

the form

- Sr , 3 3,S/)

r-0

where the first ooeft'loient of the expansion is

-1

S- tS.1

_P ',-)

y

This can be recognised as t:re soluticn of tre normal oatenary. In pract loe, the other oceffioterts

can be neglected without lcsing accuracy. _{It snould te}

noted that the series Is absolutely convergent

_{if 3}

5 1 and 2 h3. These lImIts for s cari ce acrleved as

long as the rigIdity of the riser Is relatively

_small, hioh is certainly the case for realistic values of

rlaer parameters.

Fcllowtng ?urkett's aprcaon :n whcn tre ends

'were assumed to be moment free, the solution of equation

(13) can be obtained in tre form

:1

)

_{exp-i(y-y).,(y))}

2 25/3 -

:an1(h/'/)

-mm .2 25/4 m -y ) 2 2 1/

- I

(s

-t

) s

'-yt

y; r

,2 21/i

,2j) -

7

,5 ) s

The icrient M :3 :ven oy

5

v'i :e

XLmUImCm erit crs at X

.(, /

O

- tIP ¡H

- .1 2

3:mce H T. -e ta'ie

v'i thus t.te snape o the stifer.e cs:er1ay :s given by

2

2./2 -.2

2./2.

h CI 7,, 2 2.2

-exp[-c(y-y)q,'y)}

(y

i

J)

(lia)

o ar. - hy2:112 -22 2 2 2 (1 -(h (lTh)

The tertsi:rl .dll be iverl by the sarte eXpression as -0r the :ateriary solution (See eq. 5). n oriese equations the first terms are tre soluticr.s of oatenary theory. The other terms ar:se from the ben51r

3tiftss.

For example, the angle at the subsurface

buoy, y y., rs

m-i 2 2 5/4

h

the angle at tre teck, y - y, rs

y)

tan

(h/y)

₂ 2 /3 (1g)

(h Y3

The position of maximu.m moment (-4rtich is aleo the .:west pcint) 3 given by

arid 3w - 2w, I

-- -- L

i. - -t-- 1, . s)

.2

2 2 ,s vs

it

is ¡5 nere

'2

-,

22 L-i J S -

- :h-Y3

¡ (20)

7ioat::rl

cf the

st:fferei

catenary e:uatL:s recuiras solving the set of equatCrlS (13,17) for y, ¡

ad:1 for

a jven -i3J3 of S. Tm the other rsr.d,

because o 113 :cw l:ffr.ess, the arrape of one rrser

will

be

virtually the

sarte as trat for the tatenary.

For CXample, tre nth 3 rs 200m, arti t is laXen as 503N/m-n2, tren 2 becomes .x103, 4h ion :s rtjcrl

5aller than unity.

This eens Inat rn this instance

th effect of oenclng stiffness LS very small arid

t:re catenary sol-3tions can oc used trt great accuracy.

Mcwever, Lt nust be pointed cut that tr.e stiffened

oatenary equaticns can be usefully ampio/ei in

(15) C3t imating the values for rIser tending rtomertt .ts ras teert noted, the maxirtum bending rtcment :3 311am 0/ equatIon (5).

3NAMIC AAitSS

The dynamic response of a f.ex:ble rIser :s relatively complex when compared with

its

stat::

response. This rs because cf a rr_itoer of factOrs wn:crl

include tre follcwtng

the iiffculties of defIning tre

nydrodyntainlo fortes

due to waves nd currents, wn:onl are a function cf

One dynamic respcnse

the arge ttsolacements of tre system wn:cn gIve r:ae

to non-limearities both :n terms of extreme forces (hyirodynamic) and internal forces 'tensions arti bending moments)

the existence of gecretric nor.-lir.ear:t:es irre to lar3e responses

the :r.f.ueaCe of crteX 3eneration.

Jrt view of the above, tre resulttng gcvernrn;

equations ceccme a system cf coupled non-lInear partral

differential equations. wnich can rcrmally only ce solved cv numerical means. Crie pcssibl nethcO of

solution employs the finit e element approacn, wnion wtil

be discussed later in this section. This rtethOO Is ade 00 octe woth ail the abcve factors Oh one possible exception cf Vortex ;eneratLcn ffect3.

'Iaricus

invest33tI0nS relevanttor:aertynartIs

have teert carried tut (see eferenceS 13-241, but tre

tajcrit'l

are concerned with the behavacur -cf erorer

rigid risers Cr picelines. HOwever, these pacers provide useful information cri equation formulatLcrt ant

solution 000mniques.

3y apolying NewtOn'S i_aw Irre followrg equatont can

be evel:ped for the Cyrlamic response In tr,ree

timens icns: S - -, 3v - -z-- s) -r--] vS is (13)

_Jy

ì._. ra1

'55

2 St ta

F.,

- -

r

"Q

-45L.D.D

H -

.o.C..4

and I, 4 are rS!Ctj vg velocIties _In two :rthopnsl

normal rect:ons cf e element of the riser and I and arc fluid accelerations. are components cf

of the riser in two ortno;onal normal flreCtîOls of :ne slerer.: cf ne riser (see defjnjton

il

notation).

Tre f'ni:e element crethcd is a oclvenient way to scl'e :nese ncr-linear equations. The metnod nas the

advantage that : is possible to model different parts the r?,der system

witi

tifferent

levels cf sophisticatIon. For exarpl9. areas wnere hydrcdynamic

farces are _rapidLy changing or r.ere the rIser :C:guraticn :s complicated (e.g. rear tre sub-surface :'ocy).

The

finite

clement formulation of tre dynamic

ma:r:x equat:cn in the local :ocri:r.ate system oeccmes

s.c

c.c

e e

q.} C:J _{- :K]} qJ

L L L L L. L

- ₍₂₂₎

The damping matrIx accounts for structural

damping and ta included for cmpleteness, tut generally may ce neglected oecause of tre larger hydrcdynamic dacpir.g (Fi;').

The elemental mass, stIffness and Force matr:ces

must te IsSemQLSd into a total set of equattoris in order

to arrive at a system solutIon. _{This assemoly procedure} is performed by a transformation matrix. The assembled

gIocaI system cf equattons :s then

A..A

A.A

A A A

::

cc )qJ CK3 (q) - (F) (23)

G G G C

Tre aclution for tne ncn-llnear equat:cn (23) usually caes 03e cf time dcmain solutions. The stat:c sclotion atreacy discussed ;rovtdes the

in:tlal

ocroition. Then the external loadings ant the floating

facility

motions. which are the system boundary

ocndlticns, are determined the dynamic ttme dcma:n

5Ci0tOfl can oc carri ed Out. Figure 3 shows a clock

diagram of the dynamic soLutIon cethcd. _{usually thts} requires a po'erful fIn:te element program on a Large

computer and is generally 'ery expensive. esearch is

currently underway to develop :cm.puter simulations for riser dynamics and the results of this worx 4i11

indicate the signifIcance of the dynamic response in relation to the oerfcrmance cf flexIbLe riser systems :n typical cffsncre operating Cnd survival conOitions.

.ODEL TtST3 AND RESULTS

A preliminary experimental investigation has teen carrIed out to determine the response of a typical

fexibi

riser s/stem under the action cf simulated rig motionS and envir:ren:al forces. A :31 scale model of

suoh a system was tested In a wale basin so as to

represent the tena',icur of a prototype ccmprtsing _a

semi-suomersible rig mocred In _{50m water depth and a}

single rlexltle r:ser in a typical 'lazy S'

s

358

confIguration. The test arrangement is Oepiotet , FIgure 7. The semi-submerstble rig mottOns sere

simulated using a moored surface buoy, as Shown in _the

ftgure. The riser tensions were measured using

cells located at the two ends cf

the riser.

iser rottons were obtained by analysing liceo reocrdtr.;3

o0tatned durtrg the tests.

Tests were carried out tn regular waves only, covertng normal cperattng ano siso extreme attn

:cncitions (100 year storm).

'.ode1 .Iser and Gosling

The flexicie riser was noOdled using _P.': trensOarent tube of the correctly scaled cuiStOC diameter. _{The wegrt per unit Iengtn of the riser was} adjusted to tIe approprLate value oy

ftlltr.g

It

witr

brass Drain. _{Jse of brass oran dId no: influence tre}

bencing stiffness of the model riser. An exact acalir.g

of riser flexural stiffness was not attempted for trese

preliminary tests ant would have teen almost impossible

to aohleve at the model scale adopted For these tests. Future testa will attempt to rod _{flexural st:Ffneas cy} using PIC t0oe of varying wail tnckness.

The characteristics of the moOd rtsar, at

prototype scale values, are given in Table 1.

The tests were carried cut on the casis of scaltng

to aohleve the same froude Number between tre model ano

the prototype. This meant that typioal values of Reynolds Number aonleved in tIe tests were mucn less than those for the prototype. _{In view of tre fact that} the hydrodyna!nio loading is drag-corninated,

te

inaollity to achIeve sufftciently ntgh ReyroIcs Numoer for the tests was a sgnifioant unsatsfactory feature, Future tests will te oarried Out at as arge a scale as

is feasible witn the resources and faci' _{s aiaLabLe.} Structural damping is extremely tifftcult to motel and, furthermore, prototype damping :n fcrnation is not readily available. Nowever, system damping is cominated

by rydrodynamio effots rather than structure damping.

Test Results

The model system did not exhioit ar.y Crratic

tehavtour under any of the test conditions. Typical

time nistories for the tension at the upper end of the riser are shown in figure 3A ('or art operating sea-State

(N 3.1m, T - 25) and In fIgure P.A F:r an extreme sea-state y- 16.2m, T - 12s). The _{torresoonctrg} tensIons st tne lower end of the riser are sncwn n Figures 33 anm g respectively.

The _{Ocuole amplitude of tension vartation at the}

upper end cf the riser was always less than °OS of Its mean value under the test conditions. wnion confIrms

observations reported my otners (see, for example. Reference 1).

:t

was noted Durtng the tests trat the riser

configuration often deviated signiftcantly from that

which could 5e predloted on the casis cf static theory. Thts means therefore that dynamic effects are lIkely to

be Imoortant under certain operational conditions Snd that consequently computer simulations snould tage account of these dynamic effects wherever oosstble trt the future. As would me expected tne existence of

dynamic effects would be very sensitIve to tne period of the motIon of the surface vessel and this occur red at periods less than about 7.5 secOnds under tre present test conditions. FIgure 10 shows a typical cisplacement

buoy cctior.s of

of water depth at

a wave

can be coserved that the tens

Ions at the upper

._ over ends cf the

rIser

reacn theLr peaks at

.erer.t

poInts the wave cycle. This further

Illustrates te existence

of dynamic effects in tne

::eraIi ay5tefl ceravicur.

:9

nodel

test

results houed

sclar

nrnctcr:3tiCS to trose reported by Ractdllffe in

-:ercnoe 2

:n terra of the &eneral anape cf tre tire

-t're5 for r.eer tansicr..

cc exoer:nents are :eirg planned and

testa rave already teen carred Out

n a deep water

aoxlLty oy rvsearcners at

er:Ct-att

eh:0r are

currently being evaluated.

Future tests will be carried

tut at

a

ar;er rodei scale with a view to validating

tr soculat:ons and to assesslng

the significnce

cynanLo effects vricn

are not noroally taken into

eccount in current roser analyses.

: CNS

yarine roaers are critIcally irnocrtant In

relaticn

to te design cf :'Loatlng

production systens l'or cl]. and

gas developments. t

is anticipated tnat

a nber cf

future riser systecs will be of the flexible

type and

renca partioular

attention will

need to

ce oad to

analysing the oenavour of such flexible rtser ystema.

Thos pacer ras teen concerned with desorooing cose ampler aPproaches Involved in analysIng flexible

cenavocur. Theac onclude the catenary solutions

tre aaysptctic expansion approach wnich

taes

accourt of tre roser

flextbility.

Future work will

onvclve cosplete cynaro: analysis cf

flexible rosers

sotg the fInite deceno retrtcd.

Sorne prellninary code].

tests have been carneo Out wnlch have highìignted the rportance of cynamlo affects on

the roser at certain

ove frecuencoes.

These will be the subject of future

otudy on tctn rodel expericents and nurnerical

airnulat ions.

REFERENCES

MAChADO C ano UME J M. Dynamic production riser

on Enchcva field offshore Srazil.

LatiriArerican

Dl Show Conference, Rio de Janeiro, 1-4 July iBQ.

3Ei4ET P A and FRASE J R.

Flexible riser for

a

floatong storage and offloading syates.2ffncre

Tecnrrcicgy Thr.ferer.oe, hcustcn, Texas, 'lay 19S2.

DCC 4321

NICER N N and ANCE C R. eepwater producticn

roser. ffsr.cre Technology Conference, houston,

Texas, May i33, OTC 0512.

tnglo brazilian Jenture Eyes Ocrld Floater

?rospecta. Cffsncre Engineer, January 13b5,

pp92-?ETTEI4AT 2 and UMAI J M. FlexibLe dynamic rIser

for floating production facility on the 4orth Sea.

Jffsncre Eurooe 33 Conference, Aberdeen,

S-Septernoer.

QRLEY M S. A new concept in floatIng production

cysterns. Design and Oerscional .;spects of

Floating Froduction Systeors, London, 25 depteober

364.

PASS 1. The development of a nootte :roducticn

systecir.ccroorstíng a ocapllant produotocn roser.

Secinar In Desi;n and Dperaticnal Aspects cf

Floating Production Systems, London, 25 SeptenOer

i isa.

3. Flcatong productIOn systems - the pace nots up.

Cffsrrcre Engineer, Ncvecrber 138e, pp7S-60.

g. North Sea reocrt: Loneer:ng 4or:ir Sea operator

turns rarginal prospects orto profitable producers.

Ocean :ndustry, February 955, pp35-37.

iQ Sun's versatole prcduoticn serti coobines drillIng caoaoility. oean :ndustry, February 1355,

ppCc--o.

SMOTtOLT (. Tre onfluenoe of productIon rIser design on tre corfoguration and operation cl' semi-autreraicle ficating ?roduotocn systems. ffsnore

Eurote '33 cr.ference, Aberdeen, 6-9 Septeoter

i 333.

GRIFFETMS A D. The advantages of flexiole r:aers,

Seninar or, esLgn and OperatIonal Aspecta of

Floating Proouction Systems, Lonoon, 25 Septecoer

i 93a.

O'BRll] E J, .;O'JCERT ,.<RLSTCFFERSEN 3 R

D and UiN R.

Sea test of large, nogn-pressure

flexible pipe. Offancre Technology Conference, May 19S. DCC 4739.

10. PL]JUKZTT R Static bending stresses in catenaries

and droll strings. Journal cf Engineerong for

industry, ASME. Sertes 3, ¡cl 39, Noi, Feoruary

1967, pp3l-35.

15. DEXON D A and RUT r-no D R. Stoffened cateroary

calculations

on popelone laying prctlen ...curnal

of Engoneering for Thdustry, ASME, February 1

pp153-i SO.

IS. CPARXS 2 ?. r:ne influence of tension, pressure

weignt on pope and roser defcrrnatoons and streaces.

ASME Journal of Energy Resources Tecrrnolc;y, Dol

106, March 135d, pp0550.

i 7. SAP.PI(AYA T and SAACSON M .Meo.harrics of wave

forces on cffsncre structures . ;an Nostrano

Reinnolb, New fcrx. 1931.

is.

W DAREINC and R F NEATMERS. Iarine ponelone

analysis oase :nbewt n'a IletnOd 'with an rctoo

applIcation . ournal of Engineering l'or ndustry,

Trans .&SMS, November 1370, 32(0), pp527-633.

GARDNER T 4 and FOTCH M A. Dynamic analysis of

risers and caissons

oy the Elernent Method. DTC

2551 , Offsncre Technology Conference, houston,

Texas, 975.

COWAN R and ANDRIS R P. Total pipelaying system

dynamics'. JTC 2910, Dffsnore Tecnnciogy

Conference. houston, Texas, 1977.

RATEL M h and SARCNIA S. Dynamic response of free

hanging rIsers in waves. 3rd Offshore Mechanics

and Arctic Engineerong Syrnpcs000rr. ASME, New

Orleans, Lousianna, February 1980, pp055_470.

ruM Y C and TRIANTAF'fLLOU M S. The ncnìor,esr

cynamocs of long, alencer :/loncers. 3rd Dffsnore Mecnanios ano Arctic Engineeron; Syrnposiu.m, .0311E,

New Orleans, ouisianna, February 1930, ppC77-.33.

3ERWITSP,S M M, hOFF C J and KOKARAKIS J C.

Nonlinear inverse perturlatlon in structurai

dynamics recesign of

risers.

3rd Cffsnore

Mechanics and Arctic Engineering Symposiucr, ASME, New Orleans, Louislanns, Feoruary 19614, ppbS9-437.

214. DATTA T K. Abandonment and recovery aolution of

submarine pipelines . Applied Ocean Research, 1382, y]. 4, No14. p22147-252.

25. RACTLIFFE A T. Dynamic response of flexoble

oatenary

risers.

nternatiOnal Symposium on

Developments in FloatIng ProductIon Systems,

TabLe 1. CharacteristicS cf flexible riser

bIto)

13-7/3(m)

Wejnt Lo air 2DSk

1ioiouo r.Siog raLu3 2.3ci

4cuis c 1astCiy ECO34/rrn2

n

(35 3n.i) Z3 .3 -SE BEG z I. 3 G 1 et'.

3'3ur9 : ie:<ibLe tser Svsreo -5 -i

Figure 3: Shapes oE Different Lenrh Risers

'2 __- LEFT WATER OEPTH EXCURSION POSITION lo- /

-MSTLLED POSITION I

,i

;L.

B- ___u RIGHT WATER DEPTH

5

I !//,__

EXCURSIC4 POSITION

i/I I/i

2-7/71 3

---.-.

xirnI L 8 '0 12 lU -2- ziO1 3EEE3 ,e-\

,-;-- ,,'

Fure 5: Shao. of cHe Rsers ac

ifferrc Suoy

rOS. cons

T

WATER 'VCEO RECcR TV LEVEL V IDEO CAMERA

Figure 6:

chejccc Fic'j Gcazrc fcc FecthLe Rsc

ynanic nziyz

MOL CF FLEXIBLE RISER

SUBSLFACE 3UOY CHART QECORoeR STRAJP4 GAUGE AMPtjF! ER So

Fiore 7:

odeL Tes: of che FlexibLe Riser (diensjns cn

361

sz

o,.:'-2Çc p'c?t:::z :z.s:;,s. !O,O E:c. cF FcJt :'ceo -:Ess. '"X 4ccR- .zF4 s::? = SPLAY AC

-y-M3 O°/0d Q0f d -1 3t -1 39-2 S

:re 3: ynamiC Iensic' ssc'nSeS

For r nd of Riser For Louer End of Riser

:

3 3

Figure 9: DynamiC Tension Responses

(a. ior L'por End o Rise (b) For tower End Rise: