OTC 18325 Assessment of Fixed Offshore Platform Performance in Hurricane Ivan

Puskar, F.J., Spong, R.E. and Ku, A., Energo Engineering; Gilbert, R.B. and Choi, Y.J., The University of Texas at Austin

Copyright 2006, Offshore Technology Conference This paper was prepared for presentation at the 2006 Offshore Technology Conference held in Houston, Texas, U.S.A., 14 May 2006. This paper was selected for presentation by an OTC Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Offshore Technology Conference, its officers, or members. Papers presented at OTC are subject to publication review by Sponsor Society Committees of the Offshore Technology Conference. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Offshore Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, OTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

design of fixed platforms in the Gulf of Mexico. This provides a quantified assessment of the accuracy of RP 2A and if it is adequate for design. This paper describes these assessments for Ivan based upon an in-depth study performed for the MMS focusing on fixed platforms (no caissons) [Energo Engineering, 2006]. These types of assessments have been performed previously for Andrew and Lili [Puskar, et. al., 1994 and 2004]. Recent hurricanes Katrina and Rita in 2005 provide similar opportunities, but have yet to be studied. Ivan Characteristics Ivan developed off the west coast of Africa in late August 2004. By September 5th it was a hurricane about 1,100 miles east of the southern Windward Islands. The hurricane strengthened running south of the Dominican Republic and passed within about 20 miles of Grand Cayman on the 12th. By the late afternoon on September 15th, Ivan was in the eastcentral Gulf of Mexico approaching the deepwater offshore oil and gas facilities. During this time, the hurricane was a Category 4 storm on the Saffir-Simpson scale, with maximum sustained wind speeds of more than135 mph. Figure 1 shows the storm track through the key offshore oil and gas blocks. Also shown in the figure are the fixed platforms that were destroyed during the hurricane. Ivan tracked North over the deepwater facilities in the Mississippi Canyon blocks and up into the Viosca Knoll (VK) and Main Pass (MP) block areas. The majority of the destroyed or damaged fixed platforms resided in the VK and MP block areas. Ivan continued its Northerly track through the eastern edge of the Mobile block area, making landfall as a major hurricane with maximum winds of 130 mph on the early morning of September 16th just west of Gulf Shores, Alabama. General Platform Damage A total of seven fixed platforms were destroyed in Ivan as shown in Table 1. Figure 1 shows the location of the platforms destroyed. One of the seven (MC 20A) was toppled by a mudslide, while the other six failures are thought to be attributed to metocean loads (i.e., wind, wave and current) exceeding the capacity of the structures. The seven destroyed platforms are from data provided by the MMS. Note that additional platforms may have been later decommissioned by the operator as a result of damage sustained from Ivan. There were also a number of other fixed platforms that sustained varying degrees of damage during Ivan. Some of the damage and failures were not considered a surprise, since the most of the platforms that failed or sustained major damage were older facilities. These platforms were generally

Abstract Hurricane Ivan is one of several hurricanes that have damaged or destroyed fixed offshore platforms in the Gulf of Mexico in recent years. These events provide a unique opportunity to determine the effectiveness of structural design standards and regulations and develop recommendations for changes, if needed. Specifically, Ivan provided an opportunity to evaluate the API RP 2A (RP 2A) design process for fixed platforms to ensure that it provides for well designed structures. The first part of this paper describes the general impact of Ivan on fixed platforms in terms of survival, damage or destruction. Specific findings and trends are reported related to global platform performance as well as component performance. The second part describes a quantitative assessment to determine the adequacy of the RP 2A design process. The approach uses a probabilistic based process that compares analytically predicted platform damage and survival to that actually observed during Ivan. The result is a Bias Factor that reflects how well RP 2A predicts platform behavior under hurricane loads. The work was funded by the Minerals Management Service (MMS). Introduction Ivan was one of several hurricanes in the last dozen years that have significantly damaged or destroyed fixed offshore platforms. Prior hurricanes are Andrew in 1992 and Lili in 2002. Katrina and Rita in 2005 also caused significant platform damage and destruction. These types of events provide an opportunity to determine how fixed platforms in the Gulf of Mexico perform in hurricanes on both a qualitative and quantitative basis. The qualitative basis includes a review of the typical types of damage to topsides and jacket, as well as the general trends observed, such as the number and type of platforms with wave-in-deck damage. The quantitative basis involves the comparison of the observed damage with what would have been predicted by RP 2A which is the basis for

OTC 18325

designed to lower metocean conditions and generally have lower global strength characteristics (e.g., weaker joints, less robust bracing patterns, etc.) than platforms designed and constructed to newer industry practices. Additionally, these older platforms typically have lower topside deck heights which make them significantly more susceptible to wave-indeck, which can dramatically increase the loads on the platform and cause damage. The extent of topside damage on many of the platforms, both new and older vintage, indicated Ivan caused extremely large waves and associated wave crest heights. The majority of the platforms that failed or sustained major damage during Ivan were in water depths between 200 to 350 feet and had deck heights at or below the current RP 2A minimum deck elevation requirements. The resulting damage to the topsides included deck structure failures and deformations generally as a result of wave-in-deck. Wind damage was also observed on quarters and building structures. The damage to the jackets included jacket leg buckles and separations, bracing failures (e.g., parted and buckled members), joint failures (e.g., crushed joint cans and brace punch through) and conductor bracing failures as discussed later. It is important to note that even though damage and destruction of platforms occurred during Ivan, advance warning allowed thousands of offshore workers to be safely evacuated from Gulf facilities prior to the storm reaching the area [API, Hurricane Readiness Conference, 2005]. There were also no significant environmental effects. Although a number of platforms sustained damaged, the majority of the platforms in the path of Ivan weathered the storm unscathed or with only minor damage. Figure 2 shows the percent breakdown of undamaged and damaged fixed platforms in the path of Ivan. The path represents the approximate boundaries of the hurricane strength winds, taken as 35 miles to each side of the storm centerline track, although some of the damaged platforms were outside of this path. Figure 3 shows the breakdown of the damaged and undamaged platforms based upon year installed. It is evident that older platforms sustained more damage than newer platforms. This is the same observation as for prior destructive hurricanes Andrew and Lili [Puskar, et. al., 1994 & 2004]. This clearly indicates the improvements in industry design practices with time and those newer platforms are much less susceptible to destruction and damage in hurricanes. Wave Crest and Wave Height Observations The appropriate deck height for new design as well as for structural assessments of existing platforms used in RP 2A Section 17 is a hotly debated topic since Ivan, as well as after Katrina and Rita. Figure 4 shows the deck heights of the platforms in the path of hurricane Ivan. The figure shows that the majority of platforms with decks above the RP 2A Section 2 (new design) minimum deck elevation criteria did not sustain major damage. The figure also shows a cluster of 200350 ft water depth platforms with decks lower than the RP 2A Section 2 deck criteria that either failed or sustained major damage during Ivan. One item to note in the figure is there are a number of deck heights which appear to be questionable since they are over 55 feet. The deck height data shown in the

figure was obtained from the MMS platform database. It is suspected that some of these deck heights are the cellar deck top of steel or in some cases the drill deck instead of the required cellar deck bottom of steel. Figure 5 shows the RP 2A design and Section 17 assessment wave height curves compared to the maximum wave height at the associated platform location per the Ivan hindcast [Oceanweather, 2004]. The comparison indicates the maximum wave heights during Ivan were generally in excess of the current RP 2A Section 2 wave height criteria for new design. The maximum wave heights were computed using the maximum significant wave height (Hsig) at the site during Ivan and then converting this to the maximum wave height (Hmax) as computed by the Forrestall distribution. The figure indicates that it is likely that the platforms in the path of Ivan, particularly the older platforms were exposed to loads in excess of their original design. The majority of the platforms survived due to the inherent safety factors in the designs. Additional study of the hindcast data compared to field observations indicates that several platforms had wave-in-deck damage, yet this would have not been predicted by the Ivan hindcast in terms of the crest elevation. In other words, the predicted maximum Ivan crest elevation is less than the platforms deck elevation, yet the platform sustained wave-indeck damage. This may be due to several factors including a lower hindcast Hsig than actual Hsig, uncertainties in converting Hsig to Hmax, uncertainties in computing the crest elevation from Hmax, unusual wave crest shapes, or other factors. This is a technical issue that needs to be explored further in other studies. Typical Ivan Fixed Platform Damage This section provides examples of the typical types of fixed platform damage caused by Ivan. The damage is broken down as Topsides Damage (wave-in-deck and wind induced) and Jacket Damage Topsides Damage -Wave-in-Deck The majority of the fixed platforms that sustained damage had evidence of wave-in-deck. The damage includes deflected structural members on the underside of decks, and in many cases, damage to equipment and support systems (piping, cable trays, etc.) on the lower decks. Wave inundation on older platforms with lower decks is not necessarily a surprise. However, some of the newer platforms (1990s design) also experienced wave-in-deck, although no major jacket structure damage occurred, but it did cause significant downtime and repair costs. The structural damage to the topsides consisted of distorted lower decks (plating and support under deck structure), equipment foundation deformation, and in some cases destroyed equipment shelters on the lower decks. There was also non-structural damage that was in some cases more pronounced than structural damage. This consisted of damaged equipment (power controls, generators, etc.), cable trays, and support utilities located on the lower decks of the platform. Displaced or missing grating, damaged handrails and stairs and damaged quarters hampered recovery efforts as these components need to be in good order prior to repair activities. Getting these as well as support and safety

OTC 18325

systems (power, fire water, etc.) up and running restricts the immediate and/or permanent manning of the platform, requiring work to often be done on a day-trip basis. Generally, the non-structural damage associated with wave-indeck was one of the greatest contributors of platform downtime following Ivan. Topsides Damage - Wind Wind was also a contributor to topside damage as shown in Figure 6. Numerous platforms exhibited signs of topside structural failures due to wind loading including damage to building cladding and displacement/damage of light weight structures and equipment. One of the more significant wind related issues was the damage to the temporary crew quarters on the Petronius compliant tower. The quarters and heliport toppled over into the center of the deck under wind loads, resulting in serious damage and considerable downtime [Wisch, 2005]. Jacket Damage The majority of the underwater jacket damage was confined to older platforms. Underwater jacket damage includes jacket leg failures, joint failures, brace failures and conductor guide frame failures. Examples are provided in the following. Jacket Leg Failures Local leg buckling was observed on four of the platforms that sustained major damage. Three of the SP platforms are of similar design, installed in late 1960s in approximately 230 ft water depth. All three have an 8 pile with 8 skirt pile configuration and are orientated in the same direction. Wavein-deck was observed on all three platforms and local buckles were observed on the North/Northwest legs. Ivan approached the platforms from the southeast and it was the leeward side legs under higher compression loading that buckled. A MP platform also sustained leg buckling and separation on the two diagonally opposed legs. The orientation of the platform and photos of the observed damage are shown in Figure 7. The platform is a four pile platform and the A1 and B2 legs were separated. The X-bracing was also separated at two locations near the leg damage. Note that the wave action and subsequent movement of the platform caused the leg to expand outward at the both ends, almost as if it were an external ring stiffener, as seen in Image A. Similar damage was observed in Lili [Puskar, et.al., 2004]. Joint Failures Joint failures including cracks, punching and crushing were observed on many of the platforms that sustained major damage. Several examples are shown in Figure 8. Image A shows a 24-inch diameter X-brace joint crushed under wave loading from Ivan. The platform was designed in the late 1970s. Since then, RP 2A has improved joint design formulations. In this case, a joint can (i.e., the thicker walled section of the through member) was present in the design. However, it was only marginally thicker than the connecting members and failed. Figure 8, Image B shows an example of joint punching failure. The brace was pushed through the chord member and

demonstrates a classic punching failure. This damage was located on conductor guide framing. Brace Failures The majority of the platforms with major damage sustained jacket brace failure. Most of bracing damage was local buckling of the bracing. Figure 9, Image A, shows several examples, including a buckled 24 diameter X-brace. Note that in this photo the marine growth was not cleaned off, instead it popped-off as the brace deformed. Marine growth that has popped-off in this manner is often a clue during inspections that some form of damage has occurred and further inspection is required. Figure 9, Image B, shows an example of a severed brace. The brace is 26-inch diameter x -inch wall thickness and the material yield strength is 50 ksi. Note the ends of the brace have been flattened out. This occurred after the brace separated, and then the brace ends came into repeated contact caused by the back and forth motion of the jacket during the storm. Conductor Guide Frame Failures This type of damage has been observed in Andrew and Lili and is the result of fatigue damage due to the upward and downward loads as waves pass through the structure. An example is shown in Figure 10 including the location of the conductor tray on the platform. In extreme storms like Ivan, these normally low-stress high-cycle fatigue issues become high-stress low-cycle fatigue that quickly escalates to this type of damage. The first conductor guide framing below the waterline on many platforms is between the -20 ft to -40 ft elevation and if not properly designed can be susceptible to this type of damage. The conductor tray may even come out of the water as the trough of an extreme wave like in Ivan passes the structure, and this may in addition cause wave-slam loads. The design characteristic typically causing this problem is the plating often found around the conductor guides, which dramatically increases vertical wave area/load (compared to conductor frames with tubular framing only). Another cause is the fact that many older conductor trays are attached to the legs via long-span horizontal bracing, often without any vertical bracing to the tray itself, making the tray susceptible to up-and-down vertical wave motions. This movement ultimately results in a fatigue problem. Note in Figure 10, Image A, how the steel coupon from the chord remains on the end of the separated conductor guide brace. This type of separation is characteristic of a fatigue failure where the crack initiates at the weld toe on the chord member, and over time, the crack propagates in the chord material and around the weld. Eventually the brace completely separates from the chord and may fall off of the platform. Quantitative Assessment This provides a comparison of a platforms actual response to the hurricane Ivan (destroyed, damaged or survived) versus what the load and resistance recipe in RP 2A predicts in terms of an analytical response. In other words, if a platform was destroyed in Ivan would this have been predicted by RP 2A? A probabilistic Bayesian updating process was used based upon an approach first used in 1994 for Andrew and repeated in 2004 for Lili. Details of the

OTC 18325

approach can be found in the associated references [Puskar, et. al., 1994 & 2004] and are not repeated here. The Andrew and Lili studies show that there is about 15% conservatism inherent in RP2A once all known factors of safety are removed. The approach results in what is known as a Bias Factor, which indicates the ratio of the actual capacity of the platform determined by observation to its analytical capacity determined using RP 2A. If a platform survives after a hurricane, while RP 2A analyses predicted it should have been destroyed, this platform has a Bias Factor greater than 1.0. In this case it would imply that the RP 2A analysis recipe is conservative. The Bias Factor is computed with all known factors of safety (FS) in RP 2A removed (i.e., the bias is in addition to the normal RP 2A FS). The prior work for Andrew and Lili resulted in Bias Factors of approximately 1.1 and 1.25 respectively. The bias is approximately 1.15 when Andrew and Lili are combined. These results imply that RP2A is doing a good job in terms of fixed platform design, with an inherent conservatism of about 15%. For this study, the Bias Factor was recomputed considering Ivan, based upon six platforms 2 destroyed, 3 damaged and 1 survived. The results are shown in Figure 11. The resulting Bias Factor for Ivan is 1.0, which means the prediction matches the observation almost exactly. The Bias Factor for Ivan was then combined with Andrew and Lili to determine a combined Bias Factor of 1.10 considering all three hurricanes. Note that the combined Bias Factor was calculated through a complicated probabilistic process and is not obtained by simply averaging the three individual Bias Factors. The Ivan Bias Factor is lower than for Andrew and Lili. The lower Ivan results may be explained by the particular selection of the specific platforms used to determine the Bias Factor, mostly damaged or destroyed. The inclusion of more platforms that survived Ivan, would increase the Bias Factor, but there was little information on survived platforms available to this study (most operators study damaged platforms and not those that survive). There is also a possibility that some of the damaged platforms had prior unknown existing damage that was not taken into account in the assessment. Hence the Ivan Bias Factor is believed to be conservative. Another factor for the lower Ivan Bias Factor may be the large number of wave-in-deck damaged and destroyed platforms and the associated uncertainties, such as wave crest elevation and the associated wave-in-deck loads. As noted previously, wave-in-deck issues should be investigated further. Overall, the Quantitative Assessment for Ivan indicates that the RP 2A fixed platform design approach has a Bias Factor of about 1.0. When combined with Andrew and Lili, the Bias Factor increases to 1.10. These results indicate that RP 2A is doing a conservative job. Conclusions Ivan provided an opportunity to evaluate fixed platform performance in extreme storms. The results were generally as expected, with most of the destroyed or damaged platforms of older design. As in hurricane Andrew and Lili, most of the

destroyed platforms were thought to be the result of wave-indeck loads. There were a significant amount of platforms with wavein-deck damage and wind damage that caused the platforms to be shut-in for an extended period while repairs were made. There was significantly more wave-in-deck damage for Ivan than for Lili and Andrew, indicating very large waves for Ivan. In particular, the wave crest elevations as determined from observed deck damage were exceptionally high. The quantitative assessment indicated once again that RP 2A does a good job of predicting platform performance, with a Bias Factor of 1.0 for Ivan and 1.10 (or about 10 percent conservatism) when combined with prior results for Andrew and Lili. The lower Ivan value is thought to be a combination of the platforms selected for the assessment, which were mostly destroyed or damaged and are thought to provide a conservative estimate (i.e., lower Bias Factor than actual). Acknowledgements The authors wish to thank their respective organizations for the opportunity to publish this paper. We also wish to thank the MMS for sponsoring the effort. References
ABS Consulting, Hurricane Lilis Impact on Fixed Platforms, Final Report to the Minerals Management Service, June, 2004. API, "Hurricane Readiness and Recovery Conference," Sponsored by the American Petroleum Institute, Houston, Texas, July 26-27, 2005. API, Recommended Practice for Planning, Designing and Constructing Fixed Offshore Platforms, API RP 2A, TwentyFirst Edition, 2nd Supplement American Petroleum Institute (API), Washington, D.C., October, 2005. Energo Engineering, Assessment of Fixed Offshore Platform Performance in Hurricanes Ivan Lili and Andrew, Final Report to the Minerals Management Service, Report Number 549, January, 2006. Laurendine, T. Hurricane Ivan Impact to Offshore Facilities and Status on Section 17 Assessments, Presentation given during 2005 Hurricane Readiness and Recovery Conference Production Facilities Break-out Session, Houston, Texas, July 27, 2005. MMS, Impact Assessment of Offshore Facilities from Hurricanes Katrina and Rita, News Release #3418, January 19, 2006. Oceanweather Inc., Hindcast Study of Hurricane Ivan (2004) Offshore Northern Gulf of Mexico, December 2004. OConner, P. Observations from Pompano and Nakika, Presentation given during Hurricane Readiness and Recovery Conference Production Facilities Break-out Session, Presentation on Hurricane Ivan Damage Observations, Houston, Texas, July 27, 2005. Puskar, F. J., Aggarwal, R. K., Cornell, C. A., Moses, F. and Petrauskas, C., A Comparison of Analytical Predicted Platform Damage During Hurricane Andrew, Proceedings, 26th Offshore Technology Conference, OTC No. 7473, May 1994. Puskar, F.J., Ku, A. and Sheppard, R.E., Hurricane Lilis Impact on Fixed Platforms and Calibration of Platform Performance to API RP 2A, Proceedings, Offshore Technology Conference, OTC No. 16802, May 2004. Wisch, D. Some Observations - Petronius and VK 900, Presentation given during Hurricane Readiness and Recovery Conference Production Facilities Break-out Session, Houston, Texas, July 27, 2005.

OTC 18325

No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

Area MC MP MP MP MP MP VK MP MP MP MP MP MP SP SP SP VK MP MP MP MP MP MP MP VK VK VK MP MP MP MP MP SP VK 20 98 293 293 305 306 294 296 277 279 138 311 296 62 62 62 900 281 289 290 305 305 306 306 786 780 823 261 298 144 252 280 60 989

Block A A A SONAT C E A A A B A B B A B C A A B A A B D F A-Petronius A-Spirit A-Virgo JP B-VALVE A A C D A-Pompano

Water Depth (ft) 475 79 247 232 244 255 119 212 223 290 158 250 225 340 322 325 340 307 320 289 180 241 255 271 1754 722 1130 299 222 207 277 302 193 1290

Year of Installation 1984 1985 1969 1972 1969 1969 1988 1970 2000 1998 1991 1980 1982 1967 1968 1968 1975 1999 1968 1968 1969 1969 1969 1978 2000 1998 1999 2001 1972 1968 1990 1998 1971 1994

Exposure Category L1 L1 L2 L2 L2 L2 L2 L2 L2 L2 L2 L2 L2 L2 L2 L2 L2 L2 L1 L2 L2 L2 L2 L2 L1 L1 L1 L2 L2 L2 L2 L2 L1

Deck Height Structure Type Damage Category (ft) (Note 1) 49 57.5 45 42 46 46 32 46 50.3 53.5 55 39.5 49.2 40 44 48 46.3 52 45 42 45 46 46 49 55 49 47 43 62.2 50 52 49 55.8 8-P TRI 8-P 4-P 8-P 8-P B-CAS 8-P 4-P 4-P 8-P 8-P 8-P SK 8-P SK 8-P SK 8-P 4-P 8-P 8-P 8-P 8-P 8-P 4-P SK C-TOWER 4-P OTHER 4-P 4-P 4-P SK 8-P 4-P SK destroyed destroyed destroyed destroyed destroyed destroyed destroyed major (D) major (D) major (D) major major major major major major major major major major major major major major major minor minor minor minor minor minor minor minor minor

Note 1: Damage Categories: Destroyed Complete Structural collapse of the platform. Major Severe structural overload to the primary load bearing members. Major (D) Major damage and later decommissioned. Minor Some structural damage but generally to secondary structures.
Table 1 Platforms Damaged by Hurricane Ivan

OTC 18325

: :: : :
MP 98'A' Tripod in 79" Water Depth Installed in 1985 57' Bottom Deck Height

: : : ::

: : : :: : : :

: : : : :: : :: :

: :: : :: :: :
09/16/2004 2am 87.8W 30.2N 130 mph max. wind 27.85mb pressure

:: : : :Mobile ::

: : : : :

: : : : : : :: :
Chandeleur Area, East Addition

VK 294'A' Braced Caisson in 119' Water Depth Installed in 1988 32' Bottom Deck Height

Chandeleur Area

: : : :

: : ::

: : :
Viosca Knoll

: :
S #

::: ::

: : : : :::: : : : :: : : ::: : : : : : ::::::: : : :: : : : :::: :: :

West Delta Area

: : :: : :: : : :: :: : : : :: : : : :

: : :: : : :: : :: ::

::

MC 20'A' 8-Pile in Mudslide Area 475' Water Depth Installed in 1984

: :

:: : ::

: : : :: : :: : : : : : : :: : : : :: : : # : S : Breton Sound Area : :: : : :: : :: : : : : MP 296 'A' :: : 8-Pile:in 212' Water Depth : : :: Installed in 1970 : : :: : : :: : : : MP 293'A' : : : ::::: : : : : :: :Main:Pass::Area : :::::: 8-Pile in 247' Water Depth : :: :: Installed in 1969 :: 45' Bottom Deck Height : : : : : : #: S :: : : : : ::::::: :: : : : : : :: : : :: : : : :: S: : :::::: :: : # : : :: : S # : : : : S :: : # : : MP 305'C' : ::: 8-Pile in: 244' Water Depth Installed in 1969 : S 46' Bottom # # Height Deck : :: # S SS # :: : :South Pass Area, South and East Addition South Pass Area : ::: : ::::::::::::: :: :: : : :::::::::: :: :: : South Pass Area, South and East Addition : : : S # : : :: : : : : : : :
MATTERHORN

: : :: : : : : : : :MP 293 SONAT : : : : : : : : :S SS S :# # ## S # : : : S S # #

S #
4-Pile in 232 W ater Depth Installed in 1972 42' Bottom Deck Height

: : :

: :

MP 279 'B' 4-Pile in 290' Water Depth

Main Pass Area, South and East Addition in 1998 Installed

: :

: : : : : : :

: : : :: SS ## : : :
S #

09/15/2004 10pm 88.1W 29.3N 135 mph max. winds 27.55mb pressure

: :
S #

:
S S : : #: :#:

S #
#

: :

S #

SPIRIT

S #
PETRONIUS NEPTUNE

MP 277'A' 4-Pile in 223' Water Depth Viosca Knoll Installed in 2000

S #

VIRGO

S #

:
S #

MARLIN

RAM POWELL

09/15/2004 6pm 88.2W 28.8N 135 mph max. wind 27.49mb pressure

:
MP 306'E' 8-Pile in 255' Water Depth Installed in 1969 46' Bottom Deck Height
HORN MOUNTAIN

:
Delta Area, South Addition

South Pass Area, South and East Addition

: : ::

Figure 1 Path of Hurricane Ivan Showing the Locations of the Destroyed Fixed Base Platforms.

OTC 18325

100%

90%

80% 75% Percentage of Platforms in Storm Path 70%

60%

50%

40%

30%

20% 14% 10% 6% 6%

0% Destroyed Major Damage Damage Category Minor Damage No Damage

Figure 2 - Damaged Platforms Sorted by Damage Type

80

Destroyed
70 4 4 60

Major Damage Minor Damage No Damage

Number of Platforms

50

40

One Platform Destroyed by Mudslide

30

62 3 3 1

20 4 10 11 2 0 2 Pre - 1978 1978 - 1991 (9th Edition) 1992 - 2000 (19th Edition) 23

4 2001 - Present (21st Edition)

Platform Vintage (year)

Figure 3 - Damaged Platforms Sorted by Vintage (Year Installed)

OTC 18325

80

RP 2A - Section 2 RP 2A - Section 17 - L1 RP 2A - Section 17 - L2 RP 2A - Section 17 - L3

70
MP 98 A Modified Caisson Destroyed

Destroyed Major Damage

MP 138 A 4-Pile Jacket Damage

Minor Damage No Damage

60 Deck Height (feet)

Questionable Deck Heights

50

Section 2 - L1 Section 17- L1

40
VK 294 A Braced Cassion (Designed for wave inundation) Destroyed

Section 17- L2

30

Section 17- L3

20 0 100 200 Water Depth (feet) 300 400

Figure 4 - Deck Heights of Platforms in the Path of Andrew Compared to API RP 2A Minimum Deck Elevation Requirements The deck heights were taken form operated supplied elevations to the MMS. As indicated, some of the deck elevations above 55 may be inaccurate since few platforms have decks this high.
80

70

Section 2 - L1 Section 2 - L2

Ivan Hindcast Wave Height (ft)

60

Section 17 - L1 Design Level

50

Section 17 - L2 Design Level

Section 2 - L1 Section 2 - L2 Section 17 - L1 - DL Section 17 - L2 - DL

40

30

Destroyed Platforms Major Damaged Platforms Minor Damaged Platforms

20 0 50 100 150 200 Water Depth (ft) 250 300 350 400

Figure 5 - Hindcast Maximum Wave Heights at Locations of Platforms in the Path of Andrew Compared to API RP 2A Wave Height Criteria

OTC 18325

Image B

Image C

Image A

Image D
Figure 6 - Topsides Wind Damage from Ivan Images A to D show a variety of deck equipment damaged by wind.

Platform Orientation
Leg separation A1

Platform North

A2 Leg separation

Conductors B1

B2

Leg Pile Leg

Image A

Leg

Image B

Figure 7 - Jacket Leg Damage from Ivan Image A shows the leg where it severed and was flattened due to the back and forth motion of the waves. Image B shows a severe near a circumferential weld.

10

OTC 18325

Image A

Image A

Image B

Image B

Figure 9 Brace Damage from Ivan Image A shows a buckled 24 brace. Image B shows a 26brace that completely severed.

Figure 8 - Joint Damage from Ivan Image A shows a collapsed 24 joint can that was undersized. Image B shows a brace punched completely through the chord.

OTC 18325

11

Conductors

Pump Caisson Typical boat landing structure

Waterline Elevation

Image A

Coupon remains

First elevation of conductor bracing below waterline (typically between -25 to -40 feet). This is the conductor framing elevation most susceptible to damage. Legs

Conductors Typical Fixed Platform Framing Image B Coupon remains

Figure 10 - Example of Conductor Guide Framing Damage from Ivan The figure on the right shows the typical location of conductor guide framing located near the water line that is prone to this type of damage.

Bias Factor5Comparison Figure

6.0

Combined, mean=1.10
Probability Density 5.0 4.0 3.0 Combined 2.0 1.0 0.0 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Bias Factor for Jacket Strength

Ivan, mean=1.00

Andrew, mean=1.09
Lilly, mean=1.24

Andrew

Lilly

Ivan

Figure 11 - Comparison of Bias Factors Determined for Ivan, Lili and Andrew