TO CENTRIFUGE

(1)

(2)

(3)

(4)

(5)

St. John's

CENTRIFUGE MODELLING OF LARGE SOIL DEFORMATION DUE TO ICE SCOUR

by

Paul Richard Lach, B.Se .. M.Se.

A thesi.. submiued totheSchool of Graduate Studies in partial fulfilmenlofthe requiremenlS for

the degreeof DoclOr ofPhilosophy

FacultyofEngineering and Applied Science Memorial University of Newfoundland

April 1996

Newfoundland Canada

(6)

Jlearned this at least by my experimell1 : tltar

if

aile a(/vollce,\' cOI!lit/('IIfly ill the direction of his dreams, and endeavors to live that life which he hns imagined, he will meet with a success Ilnexpected iI/ comlllOll hours. fie will put some things behilld, will pass all iI/visible bollndary; new. uuiversal, alUl more liberal laws will begin to establish ,hemstlves arouud olld wit"inllilll.

Ifyou have built castles in the air. your work need I/otbelost; ,hat is where they shouldbe. Now put the foundations Imder them.

Henry n,0reou, Walden

(7)

Abstract

kc scuurinj; IICCUl1i when drirting ice masses impinge upon and move lhrough seabed Sl.'dimcnL'i.Itis a prevalent phenomenonov~rvast areas ofthecontinental shelf regions uftheArelic Ocean al'k.lthc Canadian ellSI coaSI. This lee - seabed interaction represents 11 criticll!dcsij;oconsideration for marine pipelines associated withtheproduction of nITshorc hydrocarbon reserves. Pipelines mustbedesigned 10 accommodate loading lrnnsmiut:d by scour • induced soil deformalK>n belowthedepth oficeintrusion in a safe and cost • effective manner.Af1lional design methodology must therefore incorporate a tk.1cnninislic model which provides reliable predictions of the magninxle and extent of thesoildi~laccmentSgenel1ltedduring a scouring event.

This disscnalion describes cxperimemal aoo numerical investigations undertaken to gain a beller umlcrslanding of ice scouring effects on seabed soil.Theexperimental progrummc comprised a series of nine I:entrifuge model lests. In each test, an idealized scouring condition was simulated in an instrumented spttimcn of saturaled clay to pennil mt:asurements ofthestress and d.:fonnalion fields developedinthe:soil,thecontact pressun:sand resulLal'll folU5 acling on!hemodelicefeature,andthequalitalive effects of scour onburiedmodel pipeline segments. Test variables includedtheprescribed soil stress history, the attack angle and width of!hemodel keel,andthescourdepthattained for steady - slate conditions. Cel'llrifuge modelling was established as avaluab~tool by which10obtain insight into the mechanics of the ice • soil interaClion, and yielded quantilatlvctlataapplicable10well • defined events, Soil displacement measurements exhibited variation which wasckpendel'llupontheinitialstateofthesoil. and was also

iii

(8)

innucnced by differences in applied~tress p:llh.~ re~ulling.fmlllCh:lllg.C.~III 1l<.IUmlary conditions in individOlll scouring.eve1ll~.

The finite clement method was eVOllu:ueda~:l1\\I':;lllS I'M pn.:dicti\ll1 Ilf lhc~Ilil response under idcalizL>U scouring conditiuns. Thc aduplcd tWtl - dimensiun:!l llumerical representation incorporatc<.l a tillite strain fllnllulOltiUll and thc suil\\',l~1\\1l<.lcllcd OlS:l

two· phase nonlinear elastic· plastic material. Preliminary verificatiull Ill' the IlIll1lCric:Ll approach was provided through comparison of the analysis results with dm:! 'lC4uin.'d in representative centrifuge tests. Appropriate charactcrizatinn llf the cffccls Ill' scnuring.

required simulation of large movements associated with steady - Sl,lle cUlttlitiul\s. which imposed constraints on implementation of the analysis alld discrelizmiolltil'the suil domain. The numerical representation providcd adequate ,lppmximmiolltil"Ihe crfL'CtS nf scouring for compressible soil behaviour, where an cvcnt was char.tcterizcd hy ClllllillllUlll distonion or now, and volume change due to loss of materiOlI in the scour p:nh W:IS balanced primarily through compressive defonnation beneath the incisilln.

(9)

Acknowledgements

J havehadmany positive;,nU r\.'W3rding experiences oyertheduration of my programme llfSludic::i :tIthe Memorial University of Newfoundland. and' wish Inexp~my sincere :Iwrcciation tu fricntls. colleagues. and mentors that have supported my endeavours in a variety ClfdilTcrent ways. I have been (onunate 10beinvolvedwiththeCenu~(or Cold Ot.'CllO RcsouR:t.'S Engineering during aperiodof tremendous growth of research activity :too cllfmhility in the rltld of geotechnicalengineering.which has been initialed. inlarge measure, throughthcerrons ormysupervisor.Dr. Jack Clark. I am particularly inrlebled 10Dr. Clark for his unwavering suppon and encouragement. and his genuine interest in Illy [lCl1iOnal and professional development.

As"Iymore recently appoimed co • supervisor. Dr. Ryan Phillipshasalso had theunenviable task of wading throughthedetail of this relatively unwieldy document.

I am gr.lleful for his crealive kJeas. helpful advke. and consuuctive crilicism. all of whK:h have had considerable innuence onthedevelopmenl of this research. Dr. Phillips wasalso involved dinxtly inlheexperimental phase ofthestudy, asthe pastmanager of the Cambridge UniversilY Geolechnical CentrifugeCe~.I also wish 10 eXlend my thanks to Dr. Farrokh Poorooshasb. who introduced me 10 the: concepts of centrifuge modelling. and provided emhusiastk: guidance throughoultheearly stages.befo~his c!'Cupc 10 more lemperate climes. In addition,Jwish to thank Dr. Gary Sabin for reviewing this work. and the initial research proposal.

II was a privilege to be a visiting researcher al Cambridge Univcnity. and I am graieful 10 Professor Andrew Schofieldandthetechnical and researchstaffal lhe

(10)

Geotechnical Centrifuge Ccntre for til.: provisionl,fcxcdklll facililk."S :Ilk! in\':IIU:lhk tl.ochnicnl assistance. Dr. Colin Smith dl."Sl.:n'Cs sfll.'Ciallhankl: I'm his paliclll illqrw:tMlll on~clicalaspects of etntrifuge lesting. am his cnlighICfll.."tl advicc f1CnainingIIIdl.-sig.n and implementation of the~xp.:riments.ThectImrihulionsuftil.: hij;hlyskiliL'lJand experienced tenltifugc group. which included Messrs.Chri.~Cullisun. Slcven Ch:uk!lcr.

NeilBak~r.AdrianBrnnd.John Chandler. Paul Fonl. Ms. Sus:tn Oli"cr. andUrll.

Richard Dean and Jimmy J:lmesw~regreally appn:ciatcd. 'nlcir cheerful dispusilillll ,Ilk!

enthusiasm also crealed a Inlly enjoyable atmosphere in whichIIIwork.Iwill lllwOlyS have fond memories of the time whichIspent in England ductothe kinullcss;llll.l

hospitality which was shown to me by manyOl."Wfriend\ in the Camhridg.eSnil Mechanics group.

I was fim introduced to the phenomenon of icc Sl;OOring through par1icip;lIMIll ill field work undenaken as a par1 oftheLake Agassiz relicl 5l.our 1>1OOy inili.,ll.,j hy Dr.ChrisWoodwonh • Lynas. I am grateful to Dr. Woodworth· tynas. as a R:l,.:II:;ni1J.:,J authoriry. for passing on someofhis many insights in relation to this fascinating 3R::Il1f

~rch.More recently. I have benefited greatly from rn:qucnl Ul.'il."U.'\Sions with my colleague and friend, Dr. K.S.R. Prasad, on computational aspx:lo; involYl.,j in simulation of the physical scouring process. I also wish 10 thank Mary Booton.D.~in:'CKing. ;lnd other staff members at C • CORE forthetechnk:al alllJ mardi sUPflUrl which they provided.

I was favoured to have distinguished teachers in Drs. James Or.lImm. Len Domaschuk. and Donald Shields during my time spent as II student al the Univen;ity ul' Maniloba. I wish to extend special thanks 10themfor presentingthesubjccts or suil

vi

(11)

Jrn:dmni~s,md geUll:\;hni\;al engineering in a challenging and stimulating fashion, which enga1,!ed my imerest and ultimately influenced my \;hoscn career path.

I wish ttl acknowledge the generous financial support which I have received thmugh flust - graduate Scholarships from the Natural Sciences and Engineering Research Council of Canada, and throughitFellowship provided by the Centre for Cold Ocean Resource!; Engineering.Ialso wish to thank the Association of Canadian Universities for Northcrn Studies fur awarding tho: PolarG3.SStudemship in Engineering Research in support of this work. The necessary funding10implement the experimental study was derived in part from the Government of Canada through the Natural Sciences and Engineering Research CCluncil Strategic Grants entitledAn Integrated Investigation into /Iu:Proce.~.w!soj IceK~el/Soillmeraction, and Quonlijicotion of Seabed Damage Dlle /() IceSco/lr.

rinally. I wish10dedicate this dissertation to my parents. Christa and Edgar. for instilling in me a love of learning, and for their patience and encouragement throughout completion of the work described herein.

vii

(12)

List of Figures

1.1 500 kHlsiucscan sonar illl"!!!: (swmhwiJlh75 111)ShllWill~SCIIUl features;II15II)25 IIIwater depth interval in Resulute Bay 1.2 An umll.:rwiltcr vicwIll'ilSnl;lll icc is\;mJ keelsl:\luringinln

unconsolidated sediments

1.3 Plan and sKle elevation views of pipelineIllUVClllCllIin :, zuncIll'

large soildcfoml:ltionbt:ncalh 01 scll\!ring. icc ked 1.4 Aerial photo of the study region ncarLAlrCltc.M:llIiIUhll. Relici

iceberg scour features appcOlT 35 pmmlllcnl while lines in Cl'USS - III

cuning pallerns .. . . LJ

1.5 . Cross· seclion through large relici scour showing.marred ShC,lT pl:lOes and deformed bedding beneaththeincision surl;1cc .. . ...13 1.6 Schematic of axial section through scouring icc kl,.'Cl :llld ZlInes

ofwi!deformation as defined by Palmer 1:1 al. ( 1989) . . . 20 1.7 Soil deformation mechanisms during ice scouring. TI1P: Rupture

surface due to passive or bearing capacity failure:"ullllm :Shear dragging adjacent 10 ice keel or rupture surface . . 23 1.8 High - resolution sub - boUom prolile showing cruss - seetiunstil'

two ice scours in stratified sediment .. 21

1.9 Schematic diagram of apparent sub - scour disturbance. as exhibited on sub· bOllom profile ,,">cords . . . . 21 2.1 limiling stales of soil behaviour. shown in normaJiz....d clTectivc

stress space. .. 39

2.2 lnenial stressesin centrifuge model correspond with gr.tvity -

induced prototype stresses . 42

2.3 Variation of venital stress with depth in centrifuge model 42

2.4 Elements of the ice scour problem . . 41

2.5 Top: Multi - year pressure ridge cross - S/.'CI;on: 8011001 : Geometric model of multi - year pressure ridge . . 50

xii

(17)

2.f1 Vertical force· ddk."C1ion relationship for assumed failure

mechanism and typic<llia:condilions .53

2.1 Tup : Setlimcnl • covered nat (keel) surface of an overturned iceberg; Hollom : Ice fragmenl embedded in keberg . deformed

scahc..'tI Sl..'tIimcnts. . 57

2.K Ctlmparison bclwl'Cn basdine centrifuge model coodilions and flClddala for Beaufort Sca clays.Len :Undrained shear strength

prolill:: RighI: Overconsolidalion Ratio 67

2.9 Model iccherg geomelry and paramelers ofthescour problem 70

3.1 Sectional v)cw of laboratory consolidometer 81

3.2 Top : Schematk section ofpore pressure transducer: Bouom : Diagram of transducer insertion inlo clay specimen 89 3.3 Top: Recharging hypodennk tubing withleadsuspension:

Bollom : QUiline sketch of stages inleadtrail insertion 95 3.4 Layout ofpore pressure transducers, defonnalion markers, and

modelpipeline segments. shown in elevation and plan view 99

3.5 Sectional view of centrifuge lest package 102

3.6 Baseline model iceberg configuration and instrumentation. 105 3.7 Side elevation view of modeliceberg used in Tests01and02 . . . 106 3.8 Instrumenled model~bergsused in Tests 07and08 .110 3.9 SupponapparalUSformodelshown insideaod back

elevation views . . . 112 3.10 Plan view iIIustralton oflestpackage equipment and

instrumentation.

3.11 Components of the vane shear lest device .

. 115

..117 3.12 Idealized sectional drawing of the Cambridge Qeo(echnka!

cenlrifuge . . 119

3.13 Schematic representalion of dala acquisition system. . .122 3.14 Pre • lest vn of completed package mounted on centrifuge . . . .126 4.1 Top : Pboeographofthemodelscour crealedinTest01;Bottom :

Planview iIIustnlion of scour surface fcaNres . . . 148 xiii

(18)

4.2 Top :A\'e:ra~ccross - Sl.'Ction surrace: rR1IiIC r,lr TI,.'S1 1lI Sl:llur;

Bottom: Scour tkr1h and~mlclcvali,mn1C".t.~n:mclllS1'1111(11,.'\1 versus Ilorizomal flOSilion . . . .I~I}

4.3 Tcst 01 pore rn:ssurcI"CSfIOI\Sl$during. e:\'CIll • Ir.lll..;duccr

channels 01 1008 . 15U

4.4 Test. 01 Ilorlzomal and vl,.'11K::1ICOffipllJ'll:ntsofIhl:n:s.ultam run:c plotted againsl model horizonlal fllJSition .. 15J 4.5 Test 01 model iceberg orientalion and n:sultant fon:c \'I,.'CIIIOI . 155 4.6 Top: Plan viewdrawingshowing surface displaeemcnt VI,.'C1(lOl;

Bollom : Average axial and latcral surface displaccmcnts VCOlUS

distance from the scour axis . . 157

4.7 Test 01 initial! displaced plol of lalCral grid shown in cross -

sectionandplan view. . 151}

4.8 Test 01 axial grid at final posilioo of modcl icchcrg:Tup ;Initial

! displaced plot: Bottom: Displacement vector plot . . . ..160 4.9 Test01maximum venical and horizontal components ofsoil

displacement versusdepthbelow scourbase . . . • . .161 4. IIi Top : Plan view photogl11ph of model scour Crc:lted in Test 02:

Bottom: Plan view drawing illustrating scour surface fe:atun:s ., 166 4.11 Top : Averagecross -section surface profile forsteady -state

region of Test 02 scour; Bottom: Axialsectionplot of mcasun:d

scour depths and berm elevations 161

4.12 Test 02 port prtSSUre responses during event· tran..;duecr

channels01to 08 .169

4.13 Test 02 horizontal and venial components oftherc.~ultanlfurce plolted against model horizontal position . 171 4.14 Test 02 model iceberg orientation and resullant force VI.'CIOOl. 173 4.15 Top: Test 02 plan view drawing of surface displacement VI.'Clors;

Bouom : Average axial and lateral surface displaccmentsversu.~

distance fromthescour axis . . 115

xiv

(19)

4.16 Test 02 initialIdisplaced plot of lateral grid at x ;;; - 25 shown in cross· section and plan view . . . 177 4.17 Tcst 02 maximum venical and horizontal components of soil

displacement versus depth below scour base .. 178 4.18 Top: Phocograph of model scour created in Test 03; Bottom:

Plan view drawing illustrating scour surface features. 182 4.19 Top: Plan view photograph of model scour crealed in Test 04;

Bottom: Plan view drawing illustrating scour surface features . . . 186 4.20 Top : Average cross - section surface profile for steady - state

region of Test 04 scour; Bottom: Axial section plot of measured

scour depths and benn elevations. 187

4.21 Test 04 pore pressure responses during event - transducer

channels 01 to 08 . 189

4.22 Test 04 horizontal and venical componenlS of me resultant force

ploued against model horizontal position 191

4.23 Test 04 model iceberg orientation and resultant force vectors. .. 193 4.24 Top: Test 04 plan view drawing of surface displacement vectors;

Bottom: Average axial and lateral surface displacements versus

distance from the scour axis 194

4.25 Test 04 initialIdisplaced plot of lateral grid at x ;;; - 19 shown

in cross· sectton and plan view . 196

4.26 Test 04 initialIdisplaced plot of axial grid located at final

horizontal position of model iceberg 197

4.27 Test 04 values of vertical and horizontal components of soil displacement versusdepthbelow scour base . . . 198 4.28 Top: Plan view photograph of model scour created in Test 05;

Bottom: Plan view drawing illustrating scour surface features . . . 202 4.29 Top: Average cross • section surface profile for steady -state

region of Test OS scour; Bonom : Axial section plOI of measured

scourdepthsand henn elevations . . 203

(20)

4.30 Tes!:OSpore pressurerespo~during even! for 5Cvcn

transducer channels :!05

4.31 Test 05 horizoNal and venical componerus ofl~h.'StlIt.:I1U fon:c ploned againstmodelhorizontlll posilKln . . . 201 4.32 TestOSmodel iceberg orientation and rcsullanl forceVl.'CIOrs. " 209 4.33 Top :TestOSplanviewdrawing ofsurfitCetlisplaccmcNVl.'\.1ors:

Houom :Average axialandlaleli11 surface dispbccmcNS vcrsus

dislancefromthescour axis . . 210

4.34 TestOSinitial' displaced plot of laleral grid al x ""+41 shown

in cross·sectionand plan view 212

4.35 TestOSinitial' displaced plOi of axial grid locau.."dat final

horizontal position of model iceberg . . 213

4.36 Test 05 maximum venica!andhoriwRlal componcnts of soil

displacement versus depth below scour base 214

4.37 Top: Photograph of model scour ereated in Test 06 • view in direction of travel; Bottom: Plan view drawing illustrating scour

surface features 218

4.38 Top : Average cross . section surfaceprofilefOf steady -~ate region of Test 06 scour; Bottom: Axial sectionplotof mcasural

scourdepthsandbennelevations. . 219

4.39 Test 06pore pressure responses ploued for final secdon at reducedrate . . . .. . . 220 4.40 Test 06 horizontal and venical components oftheresultant force

plottedagainstmodelhorizontal position 222 4.41 Test 06 model iceberg orientation and resultant force vectors. .. 224 4.42 Top : Test 06 plan view drawing of surface displacemclll vectors;

Bouam :Average axial and lateral surface dispJaccmcnts versus

distance fromthescour axis . . . 226

4.43 Test 06 initialIdisplaced. plot oflateralgrid at x:Ie+54 shown in cross • section and plan view

xvi

.. 228

(21)

4.44 Tcst 06 initialJdisplaced plOI of axialaridlocated at final

borizomal position of modeliceberg .229

4.45 Test 06maximum verticalandhorizon121 components of soil

displacement versus depth below scour base 230

4.46 Top : Plan view photograph of model scour created in Test 07;

Bouom : Plan view drawing illuStrating scour surface features. . 234 4.47 Top; Avcrolge cross· section surface profile for steady. slale

region of Test 07 scour; Bouom ; Axial section plol of measured

scour deplhs and henn elevations . . 235

4.48 Test(J7pore pressure responses during event - tn.nsducc:r

channels 01 to 08 . . . . 237

4.49 Test 07 vertical component of the resultant force plaued against model horizontal position: approx. steady - scale region :

xz:: -100to.so .240

4.50 Top: Test 07 plan view drawing of surface displacement vectors:

Soltoro : Average axial and lateral surface displacements versus

distance fromthescour axis. . .. 241

4.51 Test 07 initialIdisplaced plot of lateral grid at x "" - 25 shown incross -sectton and plan view . . . 243 4.52 Test 07 initialIdisplacedplotof axialgridlocatednear

beginningorscour . . . .244

4.53 Test01 maximumvertical and horizontal componentsor soil dispbcemenlversusdepthbelowscourbase .245 4.54 Top : Plan view photograpbormodel scour created in Test 08

- view in direction of (nvcl; Bottom : Plan view drawing

iIIusualing scour surface features . . . 249

4.55 Top: Averagecross -seclion surface profik: for steady - Slate region of Test 08 scour; Bottom: Axial section plot of measured

scourdepthsand bermelevations . . 250

4.56 Test 08 pore pressureresponsesplotted for fmalsectionat

reducedrate . . • • .252

xvii

(22)

4.51 Test 08 horizontal and~nicalcompoJII.'1U oft~R.-sult:IlU ftlR."\."

plolltd againstmodel horizontalposilion 15-1

4.58 Test 08modelicebergorienl:uion and~lt:UlIfun.."1:YC.:ton 156 4059 Top : Test 08planview drawing of surface dispb..'CmCf1I\"\."\.-'tIf'S:

Bouom :Avt:ra~uialandbler.lIsurfacedispbcc:mcnlsv...-nus distance from!hescour axis _ . . . . • . • • . . . . __ • 257 4.60 TCSl 08 initial / dispbcedp10I oflaleralaridatx ". •20 sb.",,'n

in cross •sectionand planview • . . . • . . . .260 4.61 Test 08 inilial / displaetdplol:ofaxialgridIocatL"(J ncar

beginning of scour . . ...26 I

4.62 T~I08 maximum venical and horizontalcomponent~of soil

displacemcnt versus depth below scour base 262

4.63 Top: Plan view pholograph of model scour created in Tc!(t 09 BoItom : Plan view drawing illustrating scour surface fcatun:s ... 268 4.64 Top : AVCflic cross • section surface profile forSI~y- statc

region of TCSl 09 scour; Bouom : Axi.alsectionplot ofmca.'iLln:d ICOUf depthsamibermckvations . . . _ . . .269 4.65 Test 09~pressure responsesdurinae'ICN •transduc:cr

dJannds01 10 08 .. . . 270

4.66 Test 09horizonuland.verticIJcomponents of!herc:su1laftfom:

pIoCttd apiDstmodel borizonulposition ••••••.•..•...•.273 4.67 Test 09 modelk:ebc:rgoricDwionand. I'C5UItanlforce YCdDf'5 . . . . 215 4.68 Top: Test 09planview drawilll of surface disptacemc."1ll vectors:

Bouom : Avenge uial m::Ilaltral surface displacements YCfSUS

distanCe fromthescour axis . . . 276

4.69 TCSI 09 initial/displaced plot of laleralaridatx • - 40 shown

in cross· section and plan view 278

4.70 T~I09 inilial , displaced plot of axial grid localoo al final

horizontal posilion ofmodel kcbcrg . . .. 279

4.71 Test 09 maximum vertical m::I horizontal componenls of soil displacementvctSUSdepth below scour base ••.. . 280

xviii

(23)

5.1 Relationship hctween theoretical modelling and centrifuge physical

modelstudy oficcscouring . 282

5.2 Tup : Modifit.:d Cam - clay yield surface and critical state line in effective stress space: Bottom: Nonnal compression, unloading·

reloading,andcritical slate lines in compression plane . 300 5.3 (ol) Modified Cam - clay strain hardening behaviour on wet side of

critical state: predicted stress - strain response for shear test at

constoml mean effective stress . . . 304

5.3 (h) Modified Cam - clay snain softening behaviour on dry side of critical state: predicted stress - strain response for shear test at

constant mean effective stress . .305

5.4 Node ordering for interface elements and definition of rigid

surface geometry . .307

5.5 Non - local interface friction model, for which the condilion of no relative motion was approx.imated by stiff elastic behaviour, as

shown by the dashed line. .309

5.6 Two· dimensional plane strain idealization of scouring process:

finite element mesh configuration and boundary conditions adopted

in analysis . . 323

5.7 Potential soil defonnation patterns during a scouring event. Top:

Side elevation view: Bottom: View in direction of motion . . . 329 6.1 ^Test04analysis mesh configuration illustrating reference

elements andnodes .337

6.2 Top: Computed effective stress paths for reference elements ( El 20. 80.and140)during event simulation;Bottom:Element deviatoric stress againststrain . . . .338 6.3 Computed ex.cess pore pressures during event simulation

plotted against devialoric strain for reference elements ( El 20. 80.and 140) .

xix

. 339

(24)

6.4 Test 04 analysis (omourpkJls ufC;(C",'S..'pUTe pn..~oruTeltil..I for different rigid hody posiliOl'L'l. Tur :x

=

I.K Ill;1~'MlII:

x =4.lm .. J4!

6.4 Test04analysis contour riots ofCXCCS$pUTepressurc (k1':1 1 for different rigid bolIy positions.Tur : x::::0tt6 Ill: llulllIm :

x - 14.5 m.... . ]43

6.5 Test04 comparison of analysis resolts with cxrcrimcntal measurements for individualporepressure transdoccnl J.4~

6.6 Test04 analysis comact nonnalstressversus hori7.olllal ptlliililill for reference rigid surface il'l(erface dements 349 6.7 Top:Test 04comparison ofcomputt."dandmeasured rcsultalll

force componenlS; Bottom: Analysis foret: n:corcls for dillcrenl

interface friction coefficients 351

6.8 Test04analysis results· final rigid body position; Top: Displac'lJ configuration of mesh ( initial shown dashctl ); Botlom : Displace •

ment vectors ( maximum length x=14.5m ) 353

6.9 Test04analysis coruour plots ofhoritontalcomponent of suil displacement (m )for different rigid body positions.Tup :

x,. 1.8m;Bottom:x=4.1 m . . . .355

6.9 Test04 analysis comour pklts of horizontal component of soil displacement ( m ) for different rigid body positiom.Tup :

x •8.6m; Bouom : x=14.5m 356

6.10 Test04analysis contour plots ofverticalcomponent of soil displacement (m )for different rigid bodypositions.Top ; x:0.1.8 m;Bottom:x=4.1 m . . . 357 6.10 Test 04 analysis contour plots of ventcal component of soil

displacement ( m ) for different rigid body positions. Top :

x

=

8.6 m; Bottom: x • 14.5 m . . . . 358

6.11 Test04analysis contour plots displaying magnitude of plastic Slrain for different rigMt body positions. Top : x=1.8 m;

Bouom:x-4.1m . . . . 360

(25)

Cl.11 TcSI04 31litlysis contour plms displaying magnilude of plastic slrJin for diffcrcnt rigid hody posilions. Top: x ""8.6 m:

Hollum: xs:::14.5 m . . . . .361

6. 12 Tc.'il 04comparison or measured and computed values of Ihe horimnlal component of soil displacement ploned against depth

hclow the scour base 363

6.13 Tcsi04comparison of measured and computed values of the vertical component of soil displacement ploued against depth

below Ihc scour base . . . 364

Cl.14 Tesl04 comparison of computed displacement profiles before and aftcr unloading and associated elastic rebound . . . .365 6.15 Test04 analysis comparison of profiles of the horizontal displace·

ment component for different interface friction coefflcienls . . . . .367 6.16 Test 04 analysis comparison of profiles of the venical displace·

ment component for different interface friction coefficients . . . . .368 6.17 Test05I09 analysis mesh configuration illustrating reference

elementsandnodes 370

6.18 Top: Computed effective stress paths for reference elements (EL 23.97.and171 )during event simulation; Bottom: Element deviatoric stress against slrain . . . 373 6.19 Computed excess pore pressures during event simulation

plolted against deviatoric strain for reference elements

(EL 23. 97.and17( ) . . . 374

6.20 TestOS I09 analysis contour plots of excess pore pressure ( kPa ) for different rigid body positions. Top:x - 3.0m; Bottom:

X""7.1 m . . . . 376

6.20 Test05I09 analysis contour plots of excess pore pressure ( kPa ) for different rigid body positions. Top:x .,9.9m;Bottom:

X""(2.8m. . 377

6.21 Test

as

comparison of analysis results with experimental measurements for individualporepressuretransducers . . . . .379

xxi

(26)

6.22 Top : Test 05J09 3mlysis 11(lntl315ln.~s ~·vmp;ln.""wilh measuredpeak 3m!3V~I'lI.B-ev3lucs:OOllnm :Tr.msduc~r n.'Cnrd~

for inclinetl surface am!horiZOlll:l1h:I~. . J~

6.23 Tup : Test 05 / 09 comp..1rison of compulcd ,uw.l me:lsun.'d rcsllll:l111 force components; Bottom: Analysis forcer~'Curdsfur difli:rclll

interface friction coeOicicnts J~6

6.24 Tesl OSJ09analysts resuhs • final rigid hody pnsiliun: TUfl : Displaced configunlion ofmesh (initialshownda...m... \:(lolUml : Displacement vectors ( maximum Ienglh x=12.8 m) .JKK 6.25 Test 05J09analysis contour plots of horizomal L"omptlncnt of liIlil

displacement ( m ) for differenc rigid hody fIOllitions.Tul' :

x'" 3.0 m; Bottom: x

=

7.1m . "391

6.25 Test 05J093nalysis contour plocs of horizontal cumptloclM Ill" sui!

displacement ( m ) for different rigid body positions. Tup :

x - 9.9 m: Bottom: x

=

12.8 m . . .JIJ:!

6.26 Test 05J09analysis contour plots of venical tornfIOllCnc uf soil displac:emem ( m ) for differentrigKlbody positions. Tlir :

x _ 3.0m; Bottom: x=7.1 m )1))

6.26 Test 05J09analysis contour plots of venital componenttil'suil displacemenl ( m ) for different rigid botIy positions. Top:

x -9.9m; Bouom :x

=

12.8m . .JIJ4

6.27 Test 05J09 analysis contour plots displaying magnilLll.lc of plastic strain for differentrigidbodypositions.Top : )[ "" 3.0 m:

Bouam: x - 7.1 m . . . .JIJ5

6.27 Test05J09analysis contour plots displaying m:Jgnitutic ofl'laslic strain for different rigi1 body positiom. Top:x

=

^9.9^01:

Bottom:x=12.8m . . . 3%

6.28 Test OS I09comparisonofmeasurl,:(!andcompuled values Ill" the horizontal component of soil displacement plotted against depth

belowtheSCOlIrbase • . . . .400

xxii

(27)

6.29 Test 05I09 comparison of measured and computed values of the vertical cllmponent of soil displacement plotted against depth hclow the scourha~e

6.30 Maximum horizontal component of soil displacement ploued against depth below the scour base - summary of resultsfrom experimental programme .

. . . . 401

.. 417 6.31 Maximum vertical component of soil displacement ploned

against depth below the scoor base· summary of results from

experimental programme . . . 418

6.32 Schematic illustration of procedures involved in implementation of engineering model of ice· soil· pipeline interaction 425 6.33 Tt."!;t 04 comparison of predicted and measured values following

unloading. Top: Pipeline displaced configuration: Bottom:

Bending strain . . . 428

xxiii

(28)

List of Tables

1.1 Icc scourdimensions for the continent;11 shelftlfthe C;lll:ltli:1ll

Beaufort Sea

...

. . 151 . . . 131

. 142

Proposedtest matrix Test instrumentation.

Icc scour dimensions forthecastern C,'lUlli"n cuminclIl:.I

margin. . 5

Scaling relationships for centrifuge nKJdclling ·n Scaling relationships for model pipelines . . . • • . . . 81 1.2

Top; Test 01 measured pore pressureIl'llnsduccr locations:

Bollom : Maximum and minimum CXCC$ pore rrcssurcs and horizontal distance from model .

4.4 Tesc 01 resultant force data tabulated for different mtxlcl 2.1

2.2 4.1 4.2 4.3

horizontal positions -_ I~

4.5 Top : Test 02measurtd pore pressure transducerlocalion.~:

8ouom : Maximum and minimum excess poreprcs...un.-s and

horizontal distances from model 170

4.6 Test 02 resultant forcedalatabulated for different model

horizomal positions 172

4.7 Top: Test 04 measured pore pressure Inf\!iducer 100001iollS;

Bottom: Maximum and minimum excess pon: pressun..-s and horizorualdislance rrom model . . . 190 4.8 Test 04resultantroreedatatabulatedrordiffen:nc model

horizontal positions 192

4.9 Top: TestOS measured pore pressure transducerIOC3tKms;

Bottom: Maximum and minimum excess pore pressures and

horizontal distancerrom model . . . 206

4.10 Test 05 resultant roree datatabulatedror different model horizontal positions . . . 208

u:iv

(29)

4.11 TUfI : Test 06 measured porep~retransducer locations:

Ilonom : Muimumandminimum excess pon:pressures and

hnriwntal distlll1CC from model . . 221

4.12 Tc.q 06 resultant fon:c dalJ. tlIbulllted for different model

horizontal positions . . . 223

4.13 Top: Test 07 measuredpore pressure transducer locations:

Ilottom : Maximum and minimum excess pore pressures and

horizontal distance from model 238

4.14 Top: Test 08 measured pore pressure transducer locations:

Bonorn : Maximum and minimum excess pore pressures and horizowl distance from model . . . 253 4.15 TCSI 08 resultant forcedac.a13bubted for different modd

horiwntal posilions . . . . 255

4.16 Top : Tes[ 09 measured pore pressure transducer locations;

Bottom: Maximum and minimum excess pore pressures and

horizontal distance from model . 271

4.17 Test 09 resultant force data tabulated for different model horizontal positions . . . 274 5.1 Values of the interface angle of friction between kaolin and

other materials . . . • . . . 320 6.1 Summary of material pamneten for Speswhite kaolin and

additional input conditions. as specified in numeric:alanalyses . . . 335 6.2 Horizoral and venical displacement magnitudes at scour base

and limitingdepthsbelcNscourbase -summary of results from experimeral programme . . . • . . . 4l9 6.3 Model pipeline perfonnance for different initial depths of segments

below base of scour· summary of results from experimental programme . . . • . . . . 422

(30)

List of Symbols

I.ATIN

adhesion:lIt iceIsoil interface asym denotes anti· symmetric !"In of matrix

soil cohesion c.. undrained shear strength c~ coefficicOl of consolidation

c.Ix horizonaal distanc::e between model iceberg andlnlllsduLOCr"'l.~ili(ln void ratio

yield function, delineates boundary of elastically allainahlccOlllhinati\llL~

orlhc efrectivestresses acceleratkm due to gravity plastic potemial dc:pth of soil model length of drainage path

measure oftheoverdosure of imeracling surfaces k~ homol\lllipermeability coefficient

ks stiffness deruUng elastic slip at interface.pcrmilledhefon:friclitiRolIl'ilip k~ vert~1penneability coefficient

I a characteristK: length massper unitlengthof pipeline

malerial parameter in relation describing K. variation with OCR max du maximum recorded pore pressure increase

min du maximum recorded reductkm in porep~re local normal to rigid surface. tensor tjlJllntity p mean normal effective stress

p~ isotropicpreconsoJidation mean effective suess

p. equivalcrw. consolklation pressure. at currc:nt specifIC volume on l'lOl.lUflic oormal compression line

u:vi

(31)

generalized deviaLoriestress

qI dcvialoriestressal failure or maximum shearstress mial distance from centrifuge axis to point in model r.. radial disl.ancc from cenlrifuge axis10model surfa\:e sym denoles symmetric part of matrix

elapsed lime di.~placemcm,tensor quantity

uI equilibriumpore pressure at modelillsimstress Stafe velocity of movement relative to model vdocily of material point, tensor quantity soil moisture content

characteristic displacc:mem of point

horlzonl.al coordinate, position of model iceberg along scour axis curm1l spatial coordinates of material point in deformed configuration, tensor quantity

;II", eurrent coordinates of node on contact surface of mesh, tensor quantity

Xc current coordinates of rigid body reference node, tensor quanlily lateral coordinate, measured outward from scour axis

venical coordinate, with positive: values measured as depths below initial model surface

A displacement magnitude

A area of pipeline cross • section

CSL critical Slate line or locus of uttima.testaleSin effectivestressspace

o

scourdepth

o

diameter of vane blades

D rate of deformation and a measure of strain rate, tensor quantity E Young's Modulus of pipeline material

EL reference finite elemem F resultanl scour force

G elasticshearmodulusintenns of effective stresses xxvii

(32)

H height of vane blades

I momentof ineni3

I unit or kkntity matrix

IPT indul:tive pressure transduceruSl:din~(K1:SSUn:mc::ISUmnl....

J Jacobian ofelastic: deformalion - ratioormaI.."'fi31 volumeincum:nt and natura.! con(jS.UrllKm

K elastic: bulkmodulus intennsofdTCdiYl:Jlrt:SSd K. coemcienlof taleralearthpn::ssure al lUI

K.. coelTltieru of laleral canh pressure 011restin normally consolidall'd slate LVOT linearly Yillriable differential uansfonncr

Mp plaslic moment capacity N gravity and model scaling factor

N specific volume on isotropic nonnal compression line:ISl~lCimedWilh ullil value of mean erfeclive stress

NCL isotropicnormal compression line

OCRPPT overconsolidation ratio_ _

"""""=

inlenns of vertical erfcclive stresses

R rigidbodyrtlQIionat:maIerialpoint.lCDSOrquantity T peakvalueofvaneIOrqUC

T. dimeasiookss limefactor inone - dimemionalamoIidationsolution

n.c tensionloadcellusedinImriz.oruIforcerne:asurerned URL u.mo.dinc~rctoadiJllliDein awnpre:ssionplane V volume occuptedbymaterialatc:u.rrentpoint in time V. naturalreference volume or material

W buoyant weight or vertical force imposedbyIce rcatun:

GREEK

aogk of altack of k:e keel

a./3 empirically defined materialCOfISUIntsin undrained shear sll'\:nglh rclatton IS pipeline def1ection or displacement

JU(viii

(33)

angle. of (rictkID at soilIrigid surface interface

denotes virtual or infinitesimally small variation of physical measures

increment of deviatoric strain rate of slip ini .direction at contact surface

iocR:ment of volumetric strain equivalent rate of slip at contact surface

principal sirain increments

a measure of pipeline strain or relalive displacement

£, equivalent plastic strain magnitude

"-

",

oe'l

6£1·{j£~,6e,

a.

a,

o' ••

I:

f

_J plastic strain tensor

ratio of devialoric and mean erfective stresses gmlienl of unloading· reloading lines gradaentofnormal compression lines Coulomb friction coemcienl Poisson's ratio in Ienns of dfeclive stresses submerged mass density

pipeline material density a characteristic applied stress cauchy ortruestress,tensorquanlity CQructpressurestressatpoinr. on interface amenlvertaleffective suessatpointinsoilmodel preoonsoltdation vertical effectivestn:ss principal effective Slmses Kirchoff stress, tensor quantity

T", critical shear stress in Coulomb friction model

T..

equivalent shear suess at contact surface specifICvolume

xxix

(34)

u. specirlc volume on unk>ading •~lo3din~line;ls,suciilll."LIWilh unil\';II~

of mean effective stress

" effectiveangleof friclion

.<T

effetliveqle of fOOionaICOOSWlI volume

~ ~~aN~~~~~«~

angularV'Ckx:ityofcenlrifugc frequencyoflIlOIion

specific volumeon crilalSlale line associated with unitvalue ormcall effCClivestress

A denotesan incremental value

Aw central difference integration ofnileof spin,lel\!;orqu:mtit)' M critical state frictional conslanl

denotes a scalarproductof two malrices

SUBSCRIYT

devi.alOric

begiMinaoflime increment I+At end of lime incremelll

vertical SUPERSCRIPT

elastic component plastic component

(35)

Chapter 1 Introduction

1.1 Background

1.1.1 Ice Scour

Curvilinear sea bouom gougeCeatureswhich are typically one half to two metres in dtpth.lens of metres wide. and hundreds of metres or several kilometres in length arc found over vast areas oftheArctic and sub - Arctic cominental margins. These features an: allribulable 10 the process of icc scouring which occurs when drifting ice masses (icebergs, sea ice pressure rklge keels and ice island fragments) impinge upon and move through seabed sediments.Theprincipal motivation for research pertaining 10 Ihis I'henomenon arises fromthehazard posed to marine pipelines and other prospective subsea installations in cold ocean regions. Considerable literature exists regarding ice scouring. including field. experimental, and theoretical sWdics(cr.thebibliography editedbyGoodwinetal.,1985).Thedeveloping understanding of ice scouring and related design issues for marine pipelines was rcccntly summarized at an international workshop (Canada Oil and Gas Lands Adminislration, 1990).

(36)

In Nonh Amelia. comprehensive fieldstudiesof ice a:wr havebeenuI1l.k.Tt2~~n in mepolOllialproducdon areas of offshore hydrocarbon rexm:s. which includ.: the coDlinentaJ shelfrqions0(me ArcticOceanandtheCan:adi:an C1ISlcoast.Thebulk0(

available information on scour form and distribution are derived from sidesan!IiORU"

records(Frgure 1.1).includingmcs or scour cmemion evatu:ncd from n..-pctilive mapping surveys. Scour surface morphologyisabo defined baKd on high -n:soIu1ion sub •boltomprofiledataand, less rreque«I)'.through direct visual ohsctvatm during diver or manned submersible investigations (Figure 1.2).

Onthe Arcltc continenllli shelves, most scour features arcfOllm..'tIhyth~dL'CJ'I keels of sea ice pressure ridges. Average scour dimensions for lhe Canadian RcauforlSc,t are summarized in Table1.l(Lewis and BWco, 1990); however, correne dimcll'liom maydevialefrom referenced values as me f'tSlJll of ongoina revisionoflhcexistilll; llCour database.The seabedinthis rq:ionis erfectivelysaNn1cdwithIon&:.curvilinear sewn inWilerdepthsrancinlfrom 10to40 m.andrecenlornewrearu~are~n:Din WIlerdepthsupto72m.Theaverage scourisO.S mdeep(7.1m maximum)and26m wide(1375mmuimwa). ScoutlengthsareeshmaIc:dtobeontheorderofsc.eral hundredmetresto tilomeUcs.with amuimumrecordedvalueof13kin(HIWM: and Brown. 1977).Sc:ourin&taleSas high as 8.2evallSIkmIyear forthe22 to 2S mW2tet depthilllerVal.have been delenninedfrom repelitive mapping programs;although, then:

is considerable varialion inthefrequency of scour fonnation wilh changes in hUlh geographical location and water depth ontheshelf(Lewl~and Blasco, 1990).

AlongtheeaSlern Canadian seaboard, ice scouring is associated with seasonal incursions ofglacialice intheform oficebergs. SurveydatafromtheGrandBanks of

(37)

-'

-

- -~---~--- ~

". III III III I

-~

Figure 1.1 500kHzsidescan sonar image (swath width75m) showing ice scour Jeatures at15to25m water depth interval in Resolute Bay (courtesy oJ the Atlantic Geoscience Centre)

Figure 1.2Anuruienvarer view oJ a small ice island keel scouring into unconsolidated sediments (counesy of the Geological Survey oJ Canada)

(38)

Table1.1let!scol/r dimt!IU;OIU for rllt! COII/il/mllll slldl 01 tht.'GII/lldil/lllJc-dl'.fi,rf.1iI.~1 (SOIIret! :1~l\;Stl/ldBlasco. 1990)

Parameter Dimension SurVl:Y I'npul;nillll

Mean Seour DCPlh 0.5 m 10 J85 C\'ClllS

Extreme: Scour DelXh 7.im(45III water ul."f!th)

Mean Scour Width 26m 66 S49lo:vlo:nls

Extreme Scour Width 1375 m

Scour Length .5 to10km c:scimatl."tJ

Mean 8enn Width 15.3 m loolo:ven!s

Mean Oenn Height 0.7 m 100lo:venl.~

Newfoundland (Fader and King.1981:Lewis and Barrie.1981)rcve;ll a n:lativdy IIIW - density population ofmodemiceberg scoors and iceberg - cn:alt.'d SI.:abcd pit.. (Mllhil Oil Canada Ltd.,1985:Barriec:Ial..1986)at wOllerdcpIhs lesslhan about 230In(within the limit of observed iceberg dral'ls).Thecross • culling pallt.Tfl of eurvililll.:ar fumtwl'i with parallel side benns displayed in extensively scoured regions is eomll.1r.Lhle with the observed seabed morphology in the Beaufon Sea. Relic scour fnmtatKllls ..rlo:

differentiated on the basis of discordant trends, dense occurrence. andthepn:.'\Clll,:e nf fe:atures with greater dimensions. Table1.2lists characteristics for the sparse flIlfWl"linn of modem scoor featufCS onlheGrandBanksin comparison wilh surw:ydatafur olher scours on the continental shelves off Baffin Islarxl, Lahrador and Ncwfnundland, including relict features in deeper WOlters (Geonaulics ud., 1989; l..cwis lllld Blasco,1990).Themajority of measured scours within the water depth nngc nf mndern icebergdraftsarelessthan 2 mdeep,and the widths ofmostscours nOb'C hctwl.:cn 20 and 60 m, with rare0..--urrences exceeding100m (Lewis and Blasco.1990).The

(39)

Table 1.2 ICI!scour dimtnsiO/lJ for Iht taSttm Qmadian contintntal margin (suuru : GeonauticsLId.• 1989:Uwis and B/auo,1990)

Paramctl:r GrandBanks Canadianeastcoast

(modem features) (an ages)

Mean Scour Deplh 1.3m 1.6m

EXlreme Scour Depth 5.0m Il.Sm

Mean Scour Width 2' m '8m

EXlreme Scour Width 100m 330m

Scour DensilY 0.3 events'km 4 events'Ian

Water Deplh Range down⁽⁰200 m down toTSO m

Survey Population 407 events 21000evenlS

frequency of scouring on the GrandBanksisnotwell known aoo maybeellpcclcd tobe highly variable. reflecting deviations intheannual flux ofkcbergincursions. Maximum long lem scouring rales on !he northern exposed margin of the GrandBanksonthe order of 1 evenl' 100 km:' year are predicted based on available iceberg arrivaldata (Lewis and Parrott. 1987; Lewis el al .. 1988).

1.1.2 Marine Pipelines

Dn~/opml!nlScmarios

Marine pipelines offer a polentially safe. reliable. aoo cost - effective mode of conveyance for offshore hydrocarbon reserves in eold ocean regions. Pipeline lransport isnotdependent on the prevailing climatic conditions and may provide a less enviroMlCnlally lbrtalening allemalive 10 WIkertransportin ice-covered waters. To

(40)

dale. no nujor oil orgas pipelinehasbeen install'lJ in;ams whic:h an: suh';',:\:t 11\

scouringb)'seaiceoricebergs.Atprese....theHibc:mia FiddonI~GraooBanksnf NewfoundlamJistbeoN)'aaiYd)'~al'Cl underdc:vclopmc:nlforoil('WOOuc.'1ion (Bruce.1991: Chipman. 1992).Seabedsoil andicc:coodRions in Ihisn.-gionmal,;~il prohibitive(0 deliveroiland

cas

(0martelthroughmarinl: pipelines. andWltcr Iramplf1 isani...egratcomponentofthepropc:l5al productionII.-heme. 1'bcdcYelopn-ntwill includesubseaflowlinesu a part of the colleclkxl S)'S1tm toeal to theGravil)'B:uic Structure. in addition10pipelines which will expon crude oil10orrshorc It,,!ding systems situated approximately2 kilometres away from the structure.Thefuture exploilalioll of smaller fields off the Canadian east coast maybeexpected(Qincorporntc some ....;p..'\:ts of the following developmentapproaches :muhiphase pumping and flowlint:S for suhsl:a developmentS:seasonal.mobile production systems. and;icebergdetcclton. avoidance.

andmanagement prog:rammes (e.g. Lever. 1991).

In~rqions. thereispoterCiaIfor a coosidetabIeS)'Slt:mofoffshoreoil 01.00 gas transmission and p1htring pipelines (K..R. Croudale

a.

Associales. 1994).The Panarttic Drakc F - 76project:demoostJuedtbe feasibilityofproductionof asuhscaps well COIlIlICaCdtoonshoreprocc:ssfacilitieslhrouch •1200 m l10wIinebundle.from fields off Mclville: Island in the Canadian Archipelago (PalmercIal.. 1979;

Brown, 1990).On •much largerscale.Gulf Canada Resources Limitl.'CJ evaluated production fromtheAmauligak Field in the Canadian Beaufon Sea. with continuous flow to shorc via marine pipeline. overadistance of approximately 50kilomctrc.~

(Rogers. 1990). In geRenl.theinteractionof icc withtheseabed isthell'I05I.imponant design considentkm for all marine pipelinesplanned in lSS4Xiation with oil aoo gas

(41)

pnxluction in the cold ocean regions of Canada. Pipclioe design issues related to ice scouring must alsobeaddressed in regions oftheRussian Arctic where offshore dcvclopmcnLS areproposed,including areas adjacent to Sakhalin Island(e.g.Truskov arxl Surkov, 1991; Skurihin et al., 1992) and in the Kara Sea. MoS! recently, a major project hilsht:en initiated to exparxl production of onshore gas rlClds ontheYamal Peninsula in nonhem Russia. which involves the construction of six large· diameter pipelines across Baytiaratskaya Bay overa distance of approximately70kilomelres (Palmer,1994).

Pi~lill~SafelY

Theforces imposed directly ontheseabed during formation of a typical scour feature are large enough10cause severe distress toanunprotected, conventionally· designed marine pipeline, likely accompanied by rupture and loss of containment. Simplified calculations (Palmeret al.. 1990)show thailoadsamicipated during scouringeventslit:oneor[WQ

Of'dersof magnilude larger than anchor forces. which are known10causedamage.11 is not practicable to design a pipeline system which is able10withstandthelarge forces appltcd during directicccontae1;alleastnot:overconsiderabledistances alanacceptable COSI.It follows thatthepipeline mustbeprotected by burial.Theselection of a safe burial depth may delenninetheviability of pipelinetransportwhich. in some instances.

may rrnder developmcnl oftheoil orgasrlCld uneconomical. Incremental increases in the design depth bcyorxllimits achievable using conventional pipeline trenchingmethods may result in nearly exponential increases in projectedcosts(e.g. McKeehan.1990).

Inordertoevaluate theriskof direak:e •pipelmc contact associatedwitha particular burial depth. lhe following two conditionsmustberesolved : (1)the

(42)

probabililY that anicekeelwillintersect theprcscrihctJpipdinc mute durin!! tlk: tillk.' inlerval of interesl. and; (2) for a given ice ked tr:tvcBing the route. the pftlhahility that the depth of scouring will exceed the burial depth of the pif'Cline. '111e Iwer.11I risk llf failure due 10 direci contact may thenbeestahlishl.'iJ w;lh cnnsilkraliolltit'h!,lth intersection and dept.h exceedance (Comfort et aI., 1990). An eXlcn.."ivc h.1."iC!,If infonnalionhasbeen compiled on the distribultcmS of scourChar:tC1l'ri"il~'lIin tlk:

CanadianOrrShoRincluding scour depth. width. orientation, aOO l1p:llial aOO l\.'fllP.lf'il1 frequencies (e.g. Gilbertet aI., 1985; King and GiIIl.'$fl)c. 1985; Gilhcn and Pedersen, 1986; Gilbert et aI., 1989: Geonaulics LId., 1989).Thcsca~dscourdilla may be used in probabilislic analyses 10 provide a quantitative risk assessment for a prcscrihcd pipeline roule (e.g. Del noBke Verilu, 1988; Murraye::al.. 1990). Altem.1lively. icc environme~1data may be applied explicilly intheprediction of ice • pipeline enCUUfUer frequencies (e.g. PilkingtonandMarcellus, 1981; Wadhams. 1982): however, at present.

Ihis approach is Iimiled by derlCiendes intheavailable ice information. including a lack of sumcient conlinuity in coverage or resolulion of imagery or mapped data.Itm,ly he possible 10 useiceinformation 10 augmentthe seabed scour dalabase hy rdOlling thl:

differences iniceRgimcs 10 observed scour distributions. by com:lating hetW\.'Cn oclual ice features and specifIC scours or groups of scours. aOO by developing new scour ,...le predictionsbasedoniceparameters (Dickins ct aI., 1991).

An implicil assumption of early icc scour research was lh:lt a hurittl pipeline would onlybeendangered ifitwas located above(he base of the scouring ite keel. A pipeline may, however. also be damaged in a circumstance when: theicc itself did nut contact il directly.1belarge forces exerted onthe seabed duringiIscouring event mu!>l.

(43)

he IrilnSrnilleU to Ihe sui] Ilene'llh Ihc scour, inducing high stresses and possibly causing I,lrge uerormiltinns. Experiments carried Oul in a laboratory scour lank facilily IPIKlf(Klshash el al.. 1989; 1)t1orooshasb, 1989: Paulin, 1992) supponed evidence from invesligalinnsIll"relicl features and small scale field observations (Woodwonh- l.yrm,~,1990, 1992) of large sub· scour soil movements. Funhennore, in recent studies (palmer et ,II., 1989, 1990: Golder Associates l.1d., 1990), it was demonstrated that the resfllJn,'lC(If ,.pipeline subject to scouring willbe predominamly dependent on the soil uefnmlalitln intlucr.:d in the vicinity of the pipeline. The maximum stresses that canbe tmnsmiued hy the soil are limited by its strength, and, for moSI soil conditions, a typical marine pifIClinc canbeexpected to safelY resisl these stresses wilhout incurring excessive str.lins (Been, 1990). A pipeline siluated in a zone of large soil defonnalion below a scouring K:e kr..'e1 willbedeflected as a flexible structure, unless the soil is very sofl and ahle to now around the pipeline (Figure 1.3). The soil subjected to failure stresses will hemoved over large distances causing pipeline deflection and associated distress.

Thesafe burial depth of lhe pipeline must therefore be established not only below thl;: maximum depth of ice intrusion, but also beneath a zone of excessive sub - scour soil dcfonnalion. The implication of the latter condition is thai in order to facilitate rational pipeline design methods, a detenninistic model is required 10 provide reliable predictions of lhe magnitude and extent of soil defonnation during a scouring event. Once soil displacements are defined. pipeline performance canbeevaluated.Thedevelopment and verification of an appropriate model of ice • soil interaction is dependent upon the availability of quantitative dala on the effects of scouring, applicable to well - defined event.~which are relevant 10 anticipated full scale situations.

(44)

,'t'

!.

!I!::I'I,\I!"

!":I'I\',

1'"

h!! I

t t r

::!.\llllj\!.\in ::,i11

't':-t::'>-"-

IIII1UII11-11 AI I-1IIVIfll fll ('J

VERII[hl. l'IIIVI'MI HI

Figure 1.3 Pian and sidt .:In'Olion vintl.fofpi/NUni!mOwltletll illQWM(Iffllrs:t!mil dtjormlJJionMn~ha scouringicttnt

1.2 1>revlous Work

Themajority of research involving experimental and theoreticalin\lc..'''igation.~(If(he effects of icc scouringhasbeen undenaken over the last tWO dl.."ClIdes. Early labor-dlory • based studies were carried out within scour lank facilities constructed al lhe Memorial University of Newfoundland (e,g. Chari. 1975, 1979 and 1980; Chari anc.l Peters, 1981;

ChariandGrttn. 1981;Chariet al.• 1982;Grcen. 1983;Prasad.1985, and; Prasad and 10

(45)

Chari, 1986). Small scale physical model studieswere also conducted by Arctec Canada Umitcd (AhtJelnoor and LapP. 1980; Abdelnour et aI., 1981, and; Abdelnour and Grolham,1984).This work provided insight into the parameters which influence the forces applied to the seabed during a scouring event. In general, the experimental work wasusedto verify theoretical models for the prediction of the scoordepth,for given soil ronditions,of an idealizedicefeature subjtcl to specirted environmental loadings.

Early tneorettcal representations of ice scouring included the dynamic model proposed by FENCO (1975), andthework· energy models of Chari (1979) and FENCO (1975). In the dynamic model approach, the differential equation of modon was solvl.od numerically in a lime series fashion for all applied external loads, including soil, wind, wave,pack.ice.and curreN forces.Thehorizontalsoil reaction was calculated llSing Coulomb's trialwedgesolution witha plane failure surface. assuming full mohiliution of the passive pressure.Thevertical soilreactionwas caJcuJated either as a plastic material,basedontheultimate bearing capacity, or as an elastic malerial characterized by spring constanlS.Inthework· energy model approach. energy balance considerations wereusedto estimate scour depths.Theinitial kinetic energy ofthe iceberg, together with the

won:

donebyClll'Rnt -and wind • drag forces,was equated to the work expendedinicescouring of the seabed.Thewott.done attheseabedwas usedlooven:omethefully mobilized passive soilresistanceas given by CouJomb'searth pressuretheory,for an idealiud keel geometrywith a vertical scouring face.

Chari (1982) extendedthemodel by incorporating the method of stress characteristics (Sokolovski. 1965) to evaluatethesoilresistance. Comfort and Graham (1986) reviewed theavailable: theomical models of ice sc:owingandconcludedthat !beywerebestsuited

11

(46)

for the following applications: (I) assessing the risk of damage (Ihrough din-oct Clmtact) to buried pipelines byiceberg.s:(2)estimating scour pn."SSUfI:S onhurit.-dinslallalioflS.

and:(3)predicting scour potemial in;maswhereIiclddala isscarce.ThemOOcb may alsobeemployed to set physical limitalions on~ximumscourdcplhs.com..'SpORJing wilh extreme values of!heexpected driving and .esiSlive forces.

Research related to soil defomation and associated mechanisms of failure during scouring has been relatively limited.Thelimcompelling evidence oflarge sub •scour soilmovement was obtained from onshore fltld studies of relict icebergscour!>CXf!l.ISl..V on the fonner seabed of glacial Lake Agassiz in southeaslern Manitoba (Figure 1.4) (Woodworth· Lynas and Guigne.1989. 1990:WoodwOr1h·Lyna.~.1992). Detailed mapping ofthesidewalls of an excavation Ihrough a large scour approximatelySOmin width. revealed well·definedshear planes in overconsolidated clay. extending to atIca.~t 3m belowmebase of the inferred scoor incisionsurface(Figuret.S).Suh· scour displacements aslargeas3.S mWertdeduc:edfrom measured offsets of remnanl bedding.Theshearplane:s plunged atanangleofabout2Sdegreestothehoriwt1Ull whichiscomparable withtheangle of iDlCmlI friction of the Lake Agassiz clay. This evidence suggested that the soilbeneaththeice may have experienced a bearing ClpEity failure withthefonnation of a Prandtl •type mechanism. in aa:ordance with the solution of Terzaghi(1943).Clarkand PooroosJwb(1989)posculated two distinctmodesof failure during a scouring event.Thefirstmode involved the ploughing

Out

of ncar • surface material and subsequent lateral movement from the path oftheiet to btnns on bothsides oftheresulting scour incision.Thesecond modewasthat of a bearing capacicy failure. in whichthedownward movement ofthesoil wedge belowtheK:e kcc:1

12

(47)

Figure 1.4 Aerial photo of the study region near Lorette. Manitoba.

Relict iceberg scour features appear as prominent white lines in cross - cunmg panerns (couneS)' of Chris Woodwonh - Lynas,C -CORE)

.. ^,.,

Figure 1.5 Cross - .Jection through /tug/! relict scour showing mapped shear planes and deformed bedding beneath the incision surface (after Woodworth - Lynas. 1992)

13

(48)

is equivalent to the tocal volume of soil displaced from the scour lnlugh Ic.:ss the hll:!1 volume of soil comprising lhe adjacent berm S1ructures.TheinfonllationohC..ifll.'d fnml thesrudy of relict scours in Manitoba prompted Further field and Iabor-nory inv.:stiwuions to definethenature and extent of sub - scour soil deformation.

FJeld programmeswere conducted by C • CORE, St. John's. Ncwfoundlard10 investigate soil deformalion benealhmodem small scale scours formed hy p;ln icc during spring breakup onthetidalnatsofthe St.lawrence estuary near Muntmagny, Quc.ilc\:

and at Cobequid Bay. NovaScotia.Someoftheresulls of theseinvestigation~were described by PoorooshasbandClark(1990)and detailed observ3lioll!l werc prescntcd by Woodworth - Lynas(1992).The scour features studied were typically0.5to1,0OJill width,andranged in depthfromabout0.15 (00.2 m. At the Montmagny site. the scour - affecled soil wasarecently - deposited,very soft, brown sill which overlie5 a S1iffermarine clay, whereasthe CobequidBay sedimenuconsiSl~of softtofirm, highly laminated tidal silts, Excavattonswere made:throughseveral ofthescour tracb to obtolin visualrecordsoftheeffects ofscouringatthesection.and to pennit localshearS1n.-ngth measurementsusing ahand - operatedvaneshear apparatus.Theextent of .o;oil disrurbance belowthescourwasestimatedbasedon the shear S1rength measurements and also throughmappina of distoned soil horizons. This work provided evidence of sub • scour soil movements in which the pattern of soil displacements was similar to the morphology ofthescoured surface. Comouring of shear strength measurements suggcstc..'d a poorly - defined zone of slightl)' increased strength immediately below the SCllUr incision. At Cobequid Bay. denected sediment layering, small scale fokls. and faults were kJcalized benealhand immediately adjaccnc to scour features. Sedimenllayers were

14

(49)

deflected vertically l.!ownwllrd such that the magnitude of deflection diminished wim depth to negligible values approximately 0.5 m beneath me trough of a typical feature.

On bOlh edges of the trough, sediment layers were deflected upward over regions exceeding 0.8 m in widlh, correspoooing with heave ofthescoured surface. In addition.

some evidence of low angle shear plane developmeO!, similar to that observed below relict iceberg scours in Manitoba, was obtained.

Laboratory studies undenaken by C • CORE within the scour tank facilities at the Memorial University of Newfoundland included model tests conducted in both silt and sand <Poorooshasb et aI., 1989; Poorooshasb, 1989; Paulin. 1992).Thefirst set of tests investigated the scouring process in a 0.4 m thick gravity - consolidated silt unit possessing an average undrained shear strength of 4 kPa.Thesilt unit was overlain by a clay layer of 10 mm lhickness, with an average strength of 10 kPa. The entire model seabed was saturated and submerged during testing.Themodel iceberg comprised a series of aluminum plates assembled to form a complex prism shape. The model was pennitted to pitch and heave during scouring, and these movements were dependent upon the stiffness of springs which formed a part ofthemounting system.Themodel was driven forward at a constant velocity of 0.06 mls to create a scour. Tests were performed at two different scour cut depths of 40 and 70 mm.Theresults werefullyreponedby Poorooshasb et al., (1989). The effects of scouring were evaluated based on pore pressure measurements and through observation of layer defonnation. In each lest,deep • seated defonnations were observed, which appeared 10 correspond with large transverse movements in the plane perpendicular tothedirection of travel. Pore pressure measurements suggested that stress changes may be expected atdepthsk:ss than about

IS

(50)

seven times the depth of scouring, Below the smllowcr Sl.'OUr. soil displacements wcn:

recorded to a maximum dcplh or about 200 mm. wht.'n.'35 app:ln:11l n:mookJinp orthe soil. withliteobliteration orpre • exiSiing laminae. occum:d to aI.kpth"'qwl In Iwice the scour depth below the deeper feature. In gerw:ral. the effects of scouring tlt'SCrv..'tIin these experimenls tended to suppon phenomenological dalll acquired from pn:ceding liell.!

studies or relict and small scale scour features,

More rccenlly. a series of model testswere carried out to investiGale the sctlur process in both dry and submerged sand (Paulin. 1992). as a COnlinualion of the experimental programme initiated byPoorooshasb(1989).In lolal. eighl t..'SISwt.'I'C

conducted. of which two involvedsubmergedconditions.Themodelseabedwas 0.4m deep.3 m wide. and 5 m long. andwascomposed of silicasand plq)aredOIl spccmed relative densities which were varied betweenlcsts.from 01050perccnl.'Ibcmodel iceberg was constructed of aluminum plates. arranged in a regular polygonalshape.with a horizontalbase and a flat inclined front face.Thewidth of the modelwassetal either 430or860mm fora panicularItst.andtheallacle angle ofthefront face was similarly fIXed II either 15 or 30 degrees 10thehorizontal. During testing, lhe model was advanced at aconstantvelocity of 0.06 mis, Two different scour cui lk:pths of 40 and 75 mm were investigaled.

Theraults of this lest series provided fUMer indicalion of the boundary conditionsand material states for which the effects of scouring maybeexpected 10be significant.Soil displacements weremeasured10 depths below the initial surface as great as 3.5 timesthedepth of scouring. andthehorizontal component of displacement was domiDanl.Themagnitude:and extent orthesub· scour displacements increased

16