Ivan Metocean Overview

¢ Focus on deep water for now ¢ Agenda – Ivan Wind/Wave Hindcast – Current Hindcast – Wave/Wind measurements – Historical perspective – NWS Wind Forecasting ¢ Each talk followed by 5-min questions

Ivan Characteristics

Ivan… – Category 3-4 Hurricane – Central pressure 939 mb – Radius=20-30 nm

– Max Wave Hmax~96 ft – Wind=92 kt (33 (33 ft, 3030 min)

API/RP-2A 100-year…

– d/w Hmax=71.2 ft – Wind=87 kt (33 (33 ft, 3030 min)

1 Hindcast MetMethodolhodology

¢ Modeling done Modeling done by OWI ¢ Basic Steps – specify storm paramparameterseters (time history of pressure, etc.) – Run winind md modelodel to determ to determininee wind fiind field every 30eld every 30 m minutesinutes – Use modeled winds to drive wave & surge surge mmodelsodels – Validate against site measurements

Wind & nd & Wave Comparison NDBC Buoy 4204204040

2 Wind & Wave Comparisons at Marlin TLP

Wave & Wind Hindcast Summary

¢ Methods ðods & m models (Gumodels (Gumshoe)shoe) same as used for API RP2A. ¢ Excellent comparisons with Ivan measurements at buoys & platforms ¢ Gumshoe model works for Ivan ¢ , min) ¢ Hmax~96 ft; Wmax=92 kt (33’, 30 min) ¢ RP2A 100-yr: , 30 min Hmax=71 ft; W=87 kt (33’, 30 min)

3 Ivan Current Hindcast

¢ Review Hurricane Currents ¢ Hindcast Currents from Ivan ¢ Design Implications?

Hurricane Current

HurricHurricanane Cure Current: Wind Stress

¢ GenerateGenerated by d by local wlocal windind 0 stress Current Mixed Layer ¢ Strongest on right side in Momentum Transfer -50 DW (10’s of km wide)

¢ Current peaks within 1-3 De pth (m ) hours of m -100 hours of max wind Density Stratification Limits Momentum Transfer ¢ Strong inertial

-150 component persists 3-4 0 50 100 150 200 250 300 350 400 days U (cm/sec) Hurricane Current

4 Hurricane-Loop Interaction

¢ Varying temperature 0 and salinity profile -20 has strong influence -40 on hurricane cur -60 on hurricane currentrent -80 ¢ Joint hurricane-Loop -100 load cases likely De pth-120 (m) important for southern -140 -160 DW areas -180 Background TS Loop TS -200 0 50 100 150 200 250 300 U (cm/sec)

Effect of Different Temperature and Salinity on Hurricane Current profile

Hindcasting Ability… 2.0 ¢ Current hindcast 0.0 1.0

ability not as ­1.0

0 (m/sec) 0. (m/sec) developed as that fo -2.0

r Vx 0 Vy 0 -1. winds, waves -3. 0 0 -2. ¢ Little data Little data to comto compare 0 20 40 60 80 100 120 140 -4. pare Time (hours) against, no profile 0.0 data above 30 m 1.0

5 5 0. 0.

¢ Bulk mixed-layer ­ (m/sec) (m/sec) 0.0 model does good job -1.0 model ob Vx

Vy 5 0.5 in 5 of 6 comin 5 of 6 comparisons -1. 0 0 -1. with ML averages 0 20 40 60 80 100 120 140 -2. Time (hours) ML Model Compared to Measurements, Hurricane Frederic (Sept. 1979)

5 Hindcasting Ability

Hindcast Currents, Hurricane Lili ¢ Recent data shows 0

substantial shear in -20 mixed layered layer -40 ¢ M-Y 1D profile -60 model compares -80 -100 wwellell in in DDWW De pth -120 (m )

¢ Bathymetry needed -140

around shelf/slope -160 ADCP ML ¢ odels are verys are very -180 M-Y 1D ModelM HYCOM -200 sensitive to sensitive to inputs 0 50 100 150 200 U (cm/sec) Profiles Near Time of Peak Current Near Genesis

Ivan Hindcasts

¢ Commercial h hindcastindcast available with HYCOM ¢ Preliminary comparison on slope wiithth Navy Navy data shows reasonable agreement ¢ Bulk ML, M-Y 1D profile analyses also performed ¢ No DW current data for Snapshot of Ivan HYCOM Currents validation

6 Model Comparisons for Ivan in DW

Hindcast Currents, Hurricane Ivan ¢ Mixed-layer depth 0 and average speed -20 from Bulk ML, M-Y -40 1D profile models -60 similar -80 -100 ¢ HYCOM mixed- HYCOM mixed- De pth-120 (m)

layer average -140

speed is similar, -160 ML but profile and ML -180 M-Y 1D HYCOM -200 depth are 0 50 100 150 200 250 300 questionable U (cm/sec) Profiles Near Time of Peak Current in DW

Model Comparisons Continued

¢ Bulk ML, M-Y 1D Comparing ML and M­Y Models for Ivan at Fixed Latitude 250 profile model prile model prededictict M-Y ML similar currents 200 across storm track

cm/sec) 150 ¢ M-Y 1D predicts m ( M- cts 100 100 higher mean speeds U, Mean on 100 m 50

¢ HYCOM result 0 -90.5 -90.0 -89.5 -89.0 -88.5 -88.0 -87.5 -87.0 -86.5 -86.0 -85.5 includes Ulysses Longitude eddy currents, so not shown

7 Summary of Currents

¢ Ivan model efforts hampered byIvan model efforts hampered by lack of data for validation ¢ Bulk ML, M-Y 1D profile models with limited prior validation yield similar results for Iva n in DW ¢ HYCOM results inin D DW are questionable – need to investigate ¢ M-Y 1-D profile model should beD profile model should be used to derriive criteria ve criteria for shafor shallow draft platforms (Bulk ML model suitable for spars)

Industry Site Measurements at Marlin and Medusa

¢ Marlin TLP – Wind at top of crane – Wave radars on SE Marlin (noisy) ) & S& SWW sides – High sampling rates ¢ Medusa Spar – Wave radars on SE (noisy) & NW sides Medusa – High sampling rates

8 Wind Spectrumnd Spectrum at Marat Marlin

34.83.58 m/s kts a Refert 173ence fWind,t, 67.5 52.7 kts a m Elevationt 33 ft 4102.2.0 m2 /skts Reference at 170 Windft, 8, 1.51.8 m E5 kts at levation33 ft 15:30 ­ 16:30 17:30 - 18:30

2500 3500

NPD 2 3000 NPD 2000 , l , l f 2 API o freq API o re q

/Hz-s 2500 API, mid freq 2 AP I, m id fre q 2/Hz-s2 /Hz-s 2 1500 API, hi freq , m 2000 API, hi fre q m

, m Iva n at Ma rlin

i Ivan atMarlin 1500 1000 ensity ens ty, ensity D 1000 D

ergy D 500 500 En Energy Energy 0 0 0.001 0.01 0.1 1 0.001 0.01 0.1 1 Frequency, Hz Fre que ncy, Hz

1­sec gust factor = 1.36 1­sec gust factor = 1.34 1-min gust factor = 1.15 1-min gust factor = 1.14

•Gust factors agree reasonably well with NPD •Earlier spectrum agrees with NPD model but later spectrum is deficient in very low frequency energy

Wave Time Series at Marlin

Ivan Waves at Marlin

100.00 90.00 Hmax=86.3 ft 80.00 Hs,max=50.6 ft 70.00 60.00 Hs (feet)

t 50.00 Hmax 40.00 Cmax heigh 30.00 20.00 10.00 0.00 0 5 10 15 20 25 time (hours) on 15 Sep

9 Wave ave Time Series ate Series at MarMarlinlin 16:30 - 17:30 on 15on 15 Sep

Ivan Waves at Marlin, 16:30 - 17:30 on 15 Sep

60 )

(ft 40

on 20 ti 0

eleva -20

ave -40 w -60 0 600 1200 1800 2400 3000 3600 time (sec)

60 60

40 40

20 20

0 0

-20 -20

-40 -40

-60 -60 700 800 1100 1200 seconds seconds H = 73.5’ H = 86.3’

Wave Height Distribution at Marlin

10 Wave Height Distribution at Medusa

Platform Damage at Petronius and Pompano

Petronius Pompano

11 Petronius PlPlatforatform DamageDamage

Damage at 54.1’ – 57.4’ above storm water level (after accounting for 2.6’ storm surge, tide, and setdown)

Hs=51.1’ Tp=15.4s Forristall didistribution integrated overver entiree storm 100%

80%

60% hindcast ds to 60th th 3 * Correspon – 80 40% 1.0 hindcast

percentile th th damage height (40 – 70 ) percentile 20% above SWL estimate of maximum 0% 40 50 60 70 crest from Ivan hindcast max crest (feet above SWL)

Pompano Platform DamDamageage

Damage at 54.1’ – 61.3’ above storm water level (after accounting for 1.6’ storm surge and tide) Forristall distribution integrated over entire storm 100%

80%

60% hindcast 3 40% 1.0 * hindcast percentile 20% Damage height ve 0% abo SWL 40 50 60 70 max crest (m above SWL)

Corresponds to 95th – 99.9th (90th – Hs =29.2’ 99th) percentil e estimate of maximum Hs=45.4’ 60o Tp =10.2s crest from Ivan hindcast Tp=16.8s ? wave Hs=34.9’ Unlucky Crossing trains? ave nt by Tp=16.8s W enhanceme platform?

12 Wave / Platform Interaction

Model tests 80’ wave at Ekofisk

What is the height of the undisturbed wave crest (green water) that might be inferred from local platform damage?

Measurement Summary

¢ Wind spectra fit standards ¢ No evidence of “freak” (rogue) waves ¢ Distributions of measured wave crests fit design standards ¢ Damage provides no compelling evidence for criteria change

13 Ivan Characteristics

Ivan… – Pressure 939 mmb (93%) – Radius=25 nmRadius=25 nm (25%)(25%) – Forward Spd=10 kt (50%) – Wind=92 kt (33 (33 ft, 3030 min)

– Max Wave Hmax~96 ft API/RP-2A 100-year…

– d/w Hmax=71.2 ft – Wind=87 kt (33 (33 ft, 3030 min)

Key QuestionsQuestions

¢ What return interval al was was Ivan? ¢ Was Ivan van statistically “unexpected”? ¢ Should criteria be increased & if so in n what what part art of of Gulf?

Satellite image e of IIvanvan

14 Ivan’s Return Interval? Site-to-Site Variability

¢ Model hindcast (GUMSHOE) gives large site-to-site variability 100-100-yr ¢ Causes of variability 1. Water depth & fetchr depth & fetch 2. Insufficient sample oficient sample of

severe storms 96° 90 ° ° 885°5° 80° 3. Regional differences

100-yr & max Hs along the 600 ft isobath based on Gumshoe site hindcast

Insufficient Sample of Severe Storms

14 Jan 2005

60 50 Weibull fit: k = 1.20

40

30

s (ft) 40 s (ft) H H 20

10

0 20 -60 -40 -20 0 20 40 60 10 20 40 60 100 200 Dist from center (miles) Return Period

Parametric model cross-section of the Hs in Extremal fit for site near maximu m of Hurricane Camille. Camille. W/o Camille, Hs100 is 4 ft lower.

Given small size of storms & infrequent occurrence, we need several hundred years of data to sufficiently reduce this “noise”

15 Regional Variations

¢ Recent work suggests real regional variations – Severity Severity Contours – Frequency ¢ Possible physical causes – Persistent Loop water – Gulf geometry – Atmospheric influences

Storm frequency contours

Removing the “Noise”

¢ 10--44 JIP is addressing issue ¢ Solution 1: combine (pool) sites that are similar but Uncertainty of n- n-yr Hs not identical sites Model 100-yr 10k-yr Gumbel –2.5’ –5.0’ ¢ Solution 2: develop a Site deductive model Gumbel –1.6’ –2.0’ deduc model Pooled

Pooling reduces uncertainty

16 Choosing the Optimal Pooling Size

Cross-Validated Square Error for Rate Estimation ¢ Apply “cross-validation” (for omni-directional > 15mb rate) (Chouinard,1992, OTC) 0.2112 ¢ Optimal dist. ~ 100 miles 0.211 ¢ 10-4 JIP has found 0.2108 ar resu CVSE similar results 0.2106 ¢ Results that follow use 0.2104 pooling at 5-7 sites 0.2102 ¢ ill a so use t s 100 0 50 100 150 200 250 300 Will also use this 100-mi Sigma(distance) [km] scale in another key way

What Return Interval Was Ivan?

¢ ~2500 yr Hs at site where peak occurred

¢ Exceeded Hs100 over ~150 mile swath

Ivan peak vs Gumshoe N-yr Hs along 28.25°N

17 How Does Wind Compare?

¢ ~700-yr Wind Spd (33 ft, 30 min) ¢ Exceeded 100-yr over ~60 mile swath

Ivan peak vs Gumshoe N­yr Wind along 28.25°N

How Current Compare?

Ivan P e ak Curre nt vs . N-Ye ar Value s at Fixed Latitude 400 ¢ Ivan ~700-yr 50-m R=100 350 R=1000 ¢ d Exceede 100-yr 300 r ~ 20 mi s over ~ 20 mi swath 250

¢ Storm that causes 200 100-yr Hs does not 150 usually cause 100- 100

yr wind or current Uniform50 Current on U pper 50 m (cm/sec )

0 -90 -89.5 -89 -88.5 -88 -87.5 -87 -86.5 -86 -85.5 Longitude

Ivan peak vs N-yr Layered Model along 28.25°N

18 Was Ivan SStatitatistically Unexpected?

¢ Intuition: ntuition: expect one, 100-yr sstormtorm in in 100 yrs 100 yrs in entire Gulf ¢ ¢ Fact: Fact: expect Hs100 exceeded ssoomewhere here iin n Gulf every 4 yrs ¢ Because… … – Must treat GulfMust treat Gulf as stat as statististically independent independent regions – AsAsssumume “regions” in Gulfin Gulf are are 100 mi apart – Expect a 2500-yr H in 100 yrsin 100 yrs s 25 sites, ~ 100 mi apart – That sounds That sounds like Ivan!

Ivan not at a mmajorajor ssururprise e based ed on on pre-Ivan van distribution

Metocean Summary

1. Ivan generated enerated peak Hmax~ 96 ft 2. Highest in 100 n 100 yrs but notut not bby much 3. Ivan generated ~2500-yr HHs using thethe pre-Ivan extremal distribution 4. Ivan peak weak wind & ind & current ~ent ~ 700700-yr event 5. Could argue Ivan isan is an “an “outlier” 6. But newBut new desidesiggns in Eastern Gulf should include Ivan 7. Under ppeak ofeak of IvIvan, a d/w facility could have seen wave loads ~30% higher then present design but still <<< < then 100% factor ofof safsafetyety 8. Further work …… . a. Look at metocean in shallow sisitestes b. Review APIAPI metocean guidelines c. Obtain n more upper waterater--column currents

19 HURRICANE FORECASTING

Frances

Charley

Jeanne

IvanIvan

2005 Season2005 Season ForecastForecast DrDr GrGr Gray Forecast…ay Forecast…Forecast… NOAA FORNOAANOAA FOREC FORECECASTASTAST 1515 Nam15 Nam Nameed Stormed Stormd Stormsss 1212­15 Named-15 Named med StormStormss 8 88 HuHurrrricanesicanes 66­9 Hu-9 9 Hurrrricanesicanicaneses 4 44 IntenInntentense hse huse hurrurricanrricanicaneses 33­5 M-5 5 Major Huajor Hur rrriicanescanes Factors supporting an active hurricane season

•Warmer than Warmer than normal sea surface temperature •“La La Nada” •Multi-decadal AAttlantic c signal

20 2005 HURRICANE SEASON

2005 SEASON

¢ 7 Tropical Storms ¢ 2 Major Hurricanes (Dennis & Emily)

21 ACE (% of ~lea di) n Standard Deviations N N r.JI t.,,t.,,.!.t..!.t.600~~....,...., -g - g c.noc.noc.noc.noc.noc.n 0 ~ 0 ~ 1950 1951 rJJ rJJ 1953 1954 -l rJ') 1956 1957 "7= ti) 1959 1960 0 -Q. 1962 ~ =ti) .., ,,., 1963 hi Q. 1965 ,., N c 1966 =~ -·~ 1968 I a \() Q. c 1969 UI 1971 I .. =~ ~ ;- .., 1974 c. ~ 1977 ~ c ~,., =zI 1980 N = Q ;,,_ = 1983 n= L !JQ= 1986 ~ ~ ti) = ~ ti) 1989 " ~ 0 = 1992 a, > -~ =0 :P. 3 ti) -- 2001 c.= --·ti) Cl) "... 22 MAJOR HURRICANES 5 YEAR AVERAGE

23 Errors cut in half in 15 years

24 Hurricane Charley Minimum Central Pressure 9 ­ 14 August 2004 1020

1010

1000

990

(mb) 980

970 essure BEST TRACK Pr 960 Sat (TAFB) Sat (SAB)

Sat (AFWA) Charley deepened from 950 964 mb 941 Obj T­Num to mb 4 h 35 n AC (sfc) in min ear 940 l Surface landfal – NIGHTMARE! 930 8/9 8/10 8/11 8/12 8/13 8/14 8/15 8/16 Date (Month/Day)

25 FORECAST IMPROVEMENT

¢ Better Observations ¢ Improved Computer Models

How Do WDo We Track e Track A HuHurricane?rricane?

¢ Satellite IImagermagery GOES OES East East and Goes and Goes West Visual, IIR, WV Every 15-30 minutes (rapid update for r research) Used to o determine location, on, motion, and on, and intensity

¢ Aircraft Reconnaissance USAF C-130 - Primary Mission on Operations NOAA PP-3 - Primary y Mission Research NOAA NOAA G-IV – High Altitude Operations More e accurate than han satellite

¢ Doppler Ra Radardar 250 250 nm nm range for reflectivity tracking 125 125 nm nm range for Doppler velocity estimates Location, won, wind,ind, mmotiotion, rainfall estimates and and tornado ornado detection

26 High Resolution Vis

RECONNAISSANCE FLIGHT PATH

Aircraft “ALPHA” Pattern

27 NOAA G-IV AIRCRAFT

28 TRACKING AND FORECASTING HURRICANES IN 2004

Frances

Charley

Jeanne

IvanIvan

2004 Track Guidance (1st Tier)

29 2004 Track Guidance (1st Tier)

Intensity Forecasts

30 GFS TRACK FORECASTS FOR IVAN FROM 9/7/04 12Z –9/11/04 12Z HAD A SIGNIFICANT RIGHT BIAS.

31 GFS TRACK FORECASTS FOR IVAN FROM 9/13/04 12Z – 9/15/04 12Z WERE EXCELLENT IN SPECIFYING IVAN’S LANDFALL LOCATION ON GULF COAST.

IVAN CHARACTERISTICS ¢ Typical Cape Verde Storm ¢ Southern Most Major Hurricane ¢ Reached Category 5 Three Different Times ¢ Was a Category 5 for over 30 consecutive hours. ¢ Weakened and made landfall as Cat 3

32 IVAN IVAN TRACK FORECAST ERROERRORSRS

# HOURS NHC FSSE 10 YYR AV 12 24 21 44 24 47 38 78 36 79 58 112 48 108 81 146 72 161 126 217 96 222 171 248 120 289 199 319

33 WIND PROBABILITY PRODUCT

¢ Experimental Product in 2005 ¢ Available on NHC Homepage ¢ Graphical and Text ¢ Could become Operational in 2006

34 35 ¢ GGene Hene Hafele ¢ Warning and Coordination Meteorologist ¢ [email protected] ¢ 281-337-5074 x 223

36 HURRICANE-INDUCED SEAFLOOR FAILURES IN THE MISSISSIPPI DELTA by James R. Hooper, Fugro­ McClelland Marine Geosciences & Joseph N. Suhayda, Consulting Oceanographer

HURRICANE-INDUCED SEAFLOOR FAILURES IN THE MISSISSIPPI DELTA Presented to the 2005 Offshore Hurricane Readiness and Recovery Conference American Petroleum Institute Houston, Texas July 26­27, 2005

James R. Hooper, P.E. Senior Consultant Fugro-McClelland Marine Geosciences, Inc.

1 Mississippi River

New Orleans

Mississippi River Delta

THE DELTA FRONT SEAFLOOR HAS A BAD REPUTATION.

CONSIDER SHELL’S NEW SP­70B PLATFORM SHORTLY AFTER INSTALLATION IN 1969.

2 SP-7OB

Water Depth ~300 ft Seafloor Seafloor

16 Piles Driven to ~400 ft SP-70 PLATFORM 1969

SP-7OB Hurricane Camille Waves

Seafloor Seafloor Failure Zone ~100 ft Seafloor sliding

SP-70 SEAFLOOR FAILURE

3 SP-7OB Hurricane Camille Waves

Seafloor Seafloor Soil Force Against 16 Piles & 24 Well Conductors

SP-70 SOIL FORCES

SP-7OB Hurricane Camille Waves

Piles Fail at Seafloor Base of Jacket Seafloor

Soil Motion

SP-70 SOIL FORCES

4 SP-70 PLATFORM FAILURE 1969

MAJOR HURRICANE HISTORY • 9 major hurricanes (Category 3 or greater) passed near the delta 1900 - 2004. • Camille passed over the delta with a central pressure of ~909 mb. • Ivan passed east of the delta with a central pressure of ~935 mb.

5 Mobile, Alabama

New Orleans

IVAN, 2004

Mississippi River Delta CAMILLE, 1969

HURRICANE CAMILLE

• South Pass 70 landslides caused the seafloor to move downslope by more than 3000 ft in some areas. • One new 24­well platform toppled and two others were badly damaged by seafloor landslide failures. • Pipelines were also damaged and destroyed by landslides.

6 HURRICANE IVAN 2004

• Delta-front landslides during Ivan were similar in size & character to those experienced during Camille. • One platform in ~480 ft water depth toppled by landslides (MC­20). • Pipelines were also damaged and destroyed by landslides.

Western Limit of Hurricane Force Winds

Center

s

Ivan’

h of

Pat

Region of delta front discussed during this talk. From Fugro Chance, Inc. STUDY REGION PLATFORMS AND PIPELINES

7 WHY IS THE DELTA FRONT PRONE TO SEAFLOOR FAILURES SO DESTRUCTIVE TO PLATFORMS AND PIPELINES ?

Pass THE Head of Passes A’Loutre NE Pass BIRDSFOOT SE Pass

DELTA South Pass

thwest Pass

Sou

8 ~1300-1400 1840 2005 1684

18 mi 1684 1684 1840 1840 2005

2005 GROWTH OF THE DELTA

Consider this Profile line across the seafloor

GROWTH OF THE DELTA FRONT

9 From Coleman et al, 1980

Sout h Pass

Pass

Profile line

Southwest

1874 DE1979 BLTA FRONT BATHYMET ATHYM ETRY M AP RY

From Coleman et al, 1980

Sout h Pass

Pass

Mu dflow lobes Southwest

1979 DE1979 BLTA FRONT BATHYMET ATHYM ETRY M AP RY

10 Mudlobe Advance = 2.7 mi 0 1979 200 Mudlobe in 1874 1874 Mudlobe in 1979 400 Sea Level, Feet Pre-Delta (600-700 YBP)

elow 600

800

Depth B 0 2 4 6 8 10 12 Distance from South Pass, Miles

SEAWARD ADVANCE OF THE DELTA

MUDFLOWS MOVE THE DELTA FRONT SEAWARD

11 DEPOSITION IN THE DELTA

• Deposition rates 1­2 ft/yr in front of the major passes. • Sediments primarily low permeability clay with silt. • Results in weak strength profiles more than 300-ft thick, and numerous landslides that create a unique seafloor.

From Coleman et al, 1980

Region of Mudflow Gullies

Region of Mudflow Lobes SEAFLOOR GEOMORPHOLOGY

12 SEAFLOOR Mudflow Gullies or Channels GEOMORPHOLOGY

6 –10 miles

Mudflow Lobes

From Coleman et al, 1980

Slump Mudflows Mudflow Failure in the Lobe at at Head Gully End of of Gully Gully

From Coleman et al, 1980 ELONGATE LANDSLIDES

MUDFLOW ORIGIN, SEDIMENT TRANSPORT, AND DEPOSITION

13 MUDFLOW GULLIES

• Mobile sediment typically 40 to 80 feet thick. • Gully lengths up to 6 miles. • Major mudflow activity occurs several times a year in some gullies, and every few years in others.

Newest Mudflows Older Mudflows Discharging Newer Mudflows from Gullies

SEDIMENT STRENGTH PROFILES

STACKED MUDFLOW ORIGIN

14 MUDFLOW LOBES

• Individual flow thickness ranges from a few feet to ~50 feet. • Stacked mudflows are more than 100-ft thick in some areas. • Mudflow lobes tend to remain stable until triggered by large storm waves.

Mudflows in Gullies New Mudflow Lobes & Lobe Failures

Platform & Pipeline Failures

From Fugro Chance, Inc. MAP OF SEAFLOOR INSTABILITY FEATURES OFFSHORESTUDY REGION SOUTH PASS AREA

15 SW

25 ft Depth Below Sea Level, Feet 600 400 200 ST 0 0 Single f ACKED MU From 410 ft SEAFLO Col em Distance from A SOUT an, et a low 10,000 l uni , 198 0 ts H PASS AR OR PRO DFLOW 20,000 rbitrary Pre-D ~70 ft Stacked M Mudlobes Origin, Feet elta FIL 30,000 EXAMPLE Paralle EA E IN udflows l Layers 40,000 0 4 2 6

NE

Seafloor Gradient, Percent 16 Mudflow Gully Slow Feet 0 Transport Zone accumulation zone 200 Overflow Zone

Sea Level, 400 elow 600

Depth B 0 10,000 20,000 30,000 40,000 Distance from Arbitrary Origin, Feet THE FATE OF MUDFLOW SEDIMENTS

WAVE-TRIGGERED SEAFLOOR FAILURES

17 Of course, the wave is moving

Pressure increase seafloor

So the seafloor pressure Pressure is also moving Decrease

WAVE-INDUCED PRESSURE ON SEAFLOOR

Wave Wave Crest Trough

Upward Downward Motion seafloor Motion

Lateral Motion

SEAFLOOR MOTION

18 CYCLIC WAVE-BOTTOM PRESSURES • Cyclic pressure & sediment motion beneath the waves gradually weakens the sediment. • Weak sediment in the gullies slowly moves downslope during a storm. • Cyclic pressures may cause slope failure of some mudflow lobes.

SEDIMENT STRENGTH

19 Strength, KSF 0 2 4 6 0 20% of typical Feet strength Typical 200 strength in Gulf of Mexico 400 DELTA FRONT Depth Below Sea Level, 600 STRENGTH PROFILES

0 0.60.40.2 0.8 Soil Strength, KSF 0

20 B1

Feet STRENGTH 40 B2 VARIATION IN B3 B4 A SING 60 LE Seafloor, MUDFLOW

Below 80 LOBE

Depth 100

120

20 THIS STRENGTH VARIABILITY IS TYPICAL OF MUDFLOW LOBES

Soil Strength, KSF 0 0.2 0.80.60.4 1.0 0

Upper Feet 50 Bound STRENGTH PROFILES FOR

100 SLOPE STABILITY

150 Depth Below Seafloor, Lower Bound 200

21 STABILITY ANALYSES

Feet 0

200 Mudflow Lobe 300 ft WD Gully Region 400 Mudflow Lobe 400 ft WD

600

Depth Below Sea Level, 0 10,000 20,000 30,000 40,000 Distance from Arbitrary Origin, Feet

TYPICAL SEAFLOOR PROFILE

22 RESULTS OF STABILITY ANALYSES • Moderate-size waves can fail the sediments in mudflow gullies. • Mudflow Lobes with Upper-Bound strength profiles appear to be stable during intense hurricanes. • Mudflow Lobes with Lower-Bound strength profiles appear to fail during intense hurricanes.

SUMMARY

1. Sediments accumulate in shallow water and, over time, move downslope in gullies as mudflows. 2. These gully mudflows are triggered by waves typical of small to large hurricanes.

23 SUMMARY (cont.)

3. Intense hurricanes (e.g., Ivan and Camille) create large waves causing large pressures on the seafloor. 4. During Ivan/Camille-size storms, existing mudflow lobes may fail, and mudflows from gullies may overrun previously deposited mudflow lobes.

SUMMARY (cont.)

5. Large-scale seafloor failures are the primary geologic process for seaward growth of the delta. 6. Past rates of seaward growth of the delta front will likely be maintained, and seafloor failures will continue to occur.

24