You are on page 1of 12

2010 Structures Congress 2010 ASCE

2769

Life Cycle Structural Performance Assessment of Offshore Fixed Platforms


Liangsheng Wang1,
1

Fraser Munro1 and Steve Simoni1

WS Atkins, 920 Memorial City Way Suite 700, Houston, TX 77079, USA
Email: liangsheng.wang@atkinsglobal.com (L. Wang);
fraser.munro@atkinsglobal.com (F. Munro);
steve.simoni@atkinsglobal.com (S. Simoni)

ABSTRACT
Structural assessment is a key part of the structural integrity management
(SIM) process for ensuring the fitness-for-purpose of offshore fixed platforms. This
paper presents a comprehensive life cycle structural performance assessment
approach for the offshore fixed platforms. Non-linear pushover analysis is used to
perform the structural assessment, in which the platform capacity is characterized in
terms of the platforms reserve strength ratio (RSR). Comparison of the platform RSR
and the target RSR specified in the acceptance criteria determine the fitness-forpurpose of the platform. A structural reliability methodology is proposed for the
development of the regionally specific acceptance criteria. The annual probability of
platform failure is calculated using the first order reliability method (FORM). The
assessment results are used to develop a long term risk-based underwater inspection
strategy. A case study for the platforms located offshore Eastern Caribbean is used to
illustrate the application of the developed structural assessment approach.
Keywords: Structural assessment; Acceptance criteria; Fitness-for-purpose
1

INTRODUCTION

Structural Integrity Management (SIM) is an on-going process for


demonstrating the fitness-for-purpose of an offshore platform over its entire life from
installation through to decommissioning. It is an important tool for managing the
uncertainties of structural degradation, damage, changes in loading, accidental
overloading, and changes in use. Structural assessment is a key element of the overall
SIM process. The results from the assessment determine the fitness-for-purpose of the
platform and can also be used to develop an effective long-term risk-based
underwater inspection strategy.
Offshore platforms are subject to the ravages of time such as degradation,
corrosion, fatigue, seafloor subsidence, scouring, vessel impact and other accidental
damage. In addition, there is often a call to extend platform service life under revised
operating conditions such as additional wells, workover program and modifications.
The industrys evaluation of recent hurricanes in the Gulf of Mexico indicates that the
hurricane 100-year return period event is substantially higher than that presently
provided in API RP 2A 21st edition (2005). Thus, reducing the return period of the
design load results in an increase in the operating risk of offshore platforms. These

2010 Structures Congress 2010 ASCE

2770

facts underline the importance of the structural assessment and the development of
assessment acceptance criteria based on structural reliability.
Since the early 1980s there has been an increasing interest in the assessment
approaches as offshore operators worldwide became aware of the need to manage an
aging offshore platform. Industry recognized that an alternative to the traditional
component-based design checks was required in order to warrant the continued safe
operation of the offshore platforms. Significant work has been done on the
development of structural assessment methodologies. This led to the development of
assessment guidelines within API and ISO, allowing engineers to better exploit the
full capacity of structures not accounted for in the traditional design methods.
A number of investigators have discussed the structural reliability of platforms
subjected to storm wave loading. Botelho et al. (1994) calculated the failure
probability of ST130 A platform in the Gulf of Mexico during Hurricane Andrew.
Gebara et al. (2000) and Jha et al. (2000) studied the effect of sea floor subsidence on
the platform failure probability on three jacket platforms in the North Sea. HerediaZavoni et al. (2004) carried out a structural reliability assessment of deck elevations
subjected to storm wave loading for fixed platforms in the Bay of Campeche. Dier et
al. (2001) developed a reliability analysis methodology for jacket platforms in the
North Sea. Ronalds et al. (2003) investigated the structural reliability of a generic
caisson under storm overload in the Australian North West Shelf. A good collection
of recent developments in the structural reliability assessment of fixed offshore
platforms can be found in Moan (2000) and Onoufriou et al. (2001). However, to
date very limited work has been reported on the development of regionally specific
acceptance criteria which are needed to help judge whether or not a platform is fitfor-purpose.
This paper presents a comprehensive structural performance assessment
approach for offshore fixed platforms. Non-linear pushover analysis is used to
perform structural assessment, in which the platform capacity is characterized in
terms of the platforms reserve strength ratio (RSR). Comparison of the platform RSR
and the target RSR defined in the acceptance criteria determines the platforms
fitness-for-purpose. A structural reliability methodology is proposed to develop the
regionally specific assessment acceptance criteria under extreme storm events. The
annual probability of platform failure is calculated using the first order reliability
method. The acceptance criteria are developed based on platform target reliability and
the relationship between RSR and platform failure probability. The assessment results
are used to develop a long term risk-based underwater inspection strategy. Finally a
case study for the platforms located offshore Eastern Caribbean is presented to
illustrate the application of the developed structural assessment approach.
2

STRUCTURAL ASSESSMENT

The purpose of structural assessment is to establish whether an existing


structure remains fit-for-purpose or whether strengthening or repair or other
mitigation is required. Codes and standards adopt a component-based approach to
design of offshore structures. Design codes when properly applied are conservative

2010 Structures Congress 2010 ASCE

2771

and produce structures with reserves of strength. Assessment analysis seeks to utilize
the available reserve strength and redundancy not accounted for in design.
Assessment analysis can be performed by linear global analysis or non-linear
ultimate strength analysis. A linear global analysis focuses on the failure of structural
components, whereas ultimate strength analysis is a global inelastic analysis,
incorporating member yielding/buckling and joint plasticity and accounting for the
redistribution of forces to alternative load paths. An ultimate strength analysis
provides a measure of the overall strength of a platform from a system rather than
component perspective.
2.1 Structural Modeling
The structural model of a platform consists of the jacket, topsides, piles,
foundation and miscellaneous platform appurtenances, reflecting the platforms as-is
condition by accounting for known damage, deterioration and modification based on
platform inspection records.
The jacket structure is modeled as a space frame, in which tubular beam
elements are used to model the jacket structural members. For ungrouted leg-pile
annulus, the leg-pile interaction is modeled using wishbone element, which allows
only the transfer of lateral forces. For grouted piles, the leg and pile is modeled as a
composite section. Non-linear joint characteristics are included in the model to check
the capacity of each brace-chord intersection at the tubular joints. Jacket
appurtenances that are exposed to wave and current loading are included in the
model, either explicitly or implicitly by enhancing the hydrodynamic coefficients of
the primary jacket members the appurtenances are attached to. All primary members
on the topsides structures are included in the structural model to represent the correct
stiffness of the topsides. The deck girders and other primary structural components
are modeled as beam elements.
The foundation model consists of non-linear beam-column elements to
represent the piles with the soil being modeled with non-linear springs which
represent the non-linear pile-soil interaction.
2.2 Pushover Analysis
The platform ultimate strength is generally determined from a static pushover
analysis, in which environmental loads are applied progressively until the platform
collapses. Platform failure is generally defined as formation of a limiting mechanism
in the platform structure or foundation. As the load is progressively increased during
the pushover analysis, non-linear events such as member buckling and yielding, joint
plasticity, and pile pullout or plunging are monitored, which is used to detect the
formation of a limiting mechanism in the structure or foundation.
During a pushover analysis, the loadings are generally applied in two steps.
The first step consists of the platforms dead weight and topsides loads. In the second
step, the lateral environmental load is ramped up progressively until platform failure
is reached. The ramping of the environmental load should be consistent with the rate

2010 Structures Congress 2010 ASCE

2772

of increase in wave load with increasing return period and account for the point at
which wave inundation of the topsides occurs.
The platform capacity can be characterized in term of the reserve strength
ratio (RSR), which is defined as the ratio of the platforms ultimate lateral load
carrying capacity to nominal design load, normally 100-year storm lateral loading. A
target RSR is specified as the multiplier on the 100-year storm load that the platform
should be able to withstand. The platform RSR is a measure of the actual platform
capacity. Comparison of the target RSR and platform RSR values determines the
fitness- for-purpose of the structure.

Consequence of Failure

2.3 Risk Evaluation


Structural assessment can be used as a tool to develop a risk-based integrity
management strategy. The platform risk is defined as the product of the likelihood of
platform failure and the consequence of failure based on a risk matrix. The platform
RSR and the consequence of failure can also be combined to determine the overall
platform risk level. The consequence of failure considers life-safety consequence,
environmental consequence and business loss/disruption. Different operators may
have different risk matrix driven by company policy. Figure 1 presents a risk
categorization matrix example.

High

Medium
Risk

High
Risk

High
Risk

Medium

Low
Risk

Medium
Risk

High
Risk

Low

Low
Risk

Low
Risk

Medium
Risk

Low

Medium

High

Likelihood of Failure

Figure 1. Risk Matrix Example


A risk-based approach recognizes that platforms with a higher risk may
warrant more frequent and more detailed levels of inspection than platforms with a
lower risk. During the development of a risk-based underwater inspection strategy
within SIM process, the platform risk category can be used to determine inspection
intervals and work scopes, which will better focus inspection resources and optimize
inspection planning. Table 1 presents a risk-based inspection intervals recommended
in Draft API RP 2SIM.

2010 Structures Congress 2010 ASCE

2773

Risk Category

Inspection Interval Ranges

High

3-years to 5-years

Medium

6-years to 10-years

Low

11-years or greater

Table 1: Risk-Based Inspection Intervals


3

ACCEPTANCE CRITIERA

The acceptance criteria determine the minimum target RSR of the platform
required to deliver the target reliability. The target RSR can be developed based on
the platform target reliability and the relationship between the RSR and the platform
failure probability. In the following, a reliability model is introduced to deduce the
platform failure probability. This is followed by application of several approaches to
develop the target RSR.
3.1

Reliability Analysis Model

During the platform reliability analysis, a platform is modeled in terms of two


basic sub-systems: one representing jacket including foundation and the other
representing topsides. The platform fails if either the jacket sub-system or the deck
sub-system fails. The platform system reliability analysis models the two correlated
sub-systems as a system in series.
The platform failure probability is calculated for up to 8 wave approach
directions (4 orthogonal directions and 4 diagonal directions). Failure in each
direction is treated as a failure element, and the multiple direction failure forms a
series system for reliability calculations. For platforms that experience storm waves
from a dominant approach direction, probability of failure may be calculated for the
dominant direction only.
The limit state function for estimating the failure probability is defined as

g( X ) = R S = R L W

(1)

where, R is a random variable describing the platform capacity (resistance), S is the


total environmental load, L is a random variable describing the wave and current load
on the platform, and W is wind load on the platform. g(X) defines a failure criterion
that is a function of all random variables X. Failure occurs when the load S exceeds
the capacity R or when g(X) < 0.
The wave and current load L is primarily dependent on wave height and
period, wave directional spread and current profile. The current associated with a
storm wave is not typically a separate random variable. By calculating the wave and
current loads for a number of storms with different return periods, the wave and
current load can be expressed as a function of wave height using regression analysis.

2010 Structures Congress 2010 ASCE

2774

The wave and current load characteristics change when the wave crest reaches
the bottom of steel of the lowest deck of the topsides. For the regression of wave and
current load as a function of wave height, two different functional forms are selected
for wave and current load before and after deck inundation. The wave and current
load L has the functional form

L = C1 H mC2

C2
L = [C1 + C 3 (H m H d )] H m

Hm < Hd
Hm Hd

(2)

where, Hd is the maximum wave height which results in a wave crest just hitting the
underside of lowest deck of topsides, Hm is a random variable describing the annual
maximum wave height, C1 is a random variable that characterizes the load uncertainty
on the jacket system, C2 is the exponent in the load-wave height relationship, and C3
is a random variable that characterizes the uncertainty in the wave-in-deck loads.
With the establishment of a limit state function, the annual probability of
failure is defined by the expression:

Pf = (1 FS ( x) ) f R ( x) dx
0

(3)

where, Pf is the annual probability of failure, FS is the cumulative distribution


function (CDF) for the environmental load S, and fR is the probability density function
(PDF) for the capacity R.
The probability of failure can be calculated using the Rackwitz-Fiessler
FORM method. In this method, the probability of failure is estimated using first
order approximation to the limit state at the design point. Non-normal distribution
random variables are transformed to equivalent normal distribution variables. The
Newton-type recursive formula is used to iteratively search the most probable point
(MPP) of failure in the limit state surface. The probability of failure is calculated as
Pf = ( ) , where is reliability index and () is the cumulative distribution
function of a standard normal variable.
3.2

Acceptance Criteria

For the development of regional acceptance criteria, a number of platforms are


chosen as representatives. The RSR for each platform is calculated from the pushover
analyses. The failure probability is also calculated from the reliability analyses. The
relationship between the RSR and probability of failure is hence established. Then the
target RSR can be developed based on the target reliability, i.e. acceptable failure
probability, and the relationship between RSR and platform probability of failure.
Several approaches have been investigated in the literature to develop the target
reliability, such as the method based on progressive collapse limit state, the method
based on fatal accidental rate (FAR) and the method based on ALARP principle (as
low as reasonably practicable). Some operators also refer to the recommended or
implied target reliability in offshore assessment practices in other parts of the world.

2010 Structures Congress 2010 ASCE

2775

The approach to development of target reliability based on the progressive


collapse limit state is based on meeting a 10,000-yr ultimate load return period. The
key point is that only the environmental event uncertainty is specified to achieve the
0.0001 probability of occurrence of the extreme storm event. Consequently, the
capacity and loading uncertainties are effectively set to zero (Stahl et al. (2000)).
FAR has been used to develop the target reliability. FAR is defined as
FAR = ( PLL 10 8 ) /( POB T ) , where PLL is the potential loss of life, POB is the
number of person on board, and T is the number of hours of exposure per year. The
limit of FAR value for the offshore platform can be referred to HSE report OTR
2001-63 (2002). The actual FAR value for the extreme storms can be derived based
on a quantitative risk assessment (QRA). Then the acceptable probability of failure
can be calculated as Pf = PLL / POB .
Recently ALARP principle has been used for the development of target
reliability using ICAF (implied cost to avert a fatality). This approach seeks the
balance between costs and risks in making the choice of a failure probability limit.
4

CASE STUDY

A case study for the platforms located offshore Eastern Caribbean is presented
to illustrate the application of the structural assessment approach.
4.1

Platform Assessment
Two representative platforms, i.e., Platform A and Platform B were studied.
Platform A is an 8-leg 8-pile fixed steel jacket platform operating in 67 meter of
water, which represents a more-robust structure. Platform B is a 4-leg 4-pile fixed
steel jacket platform operating in 72 meter of water, which represents a less-robust
structure. Three-dimensional structural models have been created in USFOS program
for these two platforms based on as-built drawings. The as-is platform conditions
were represented by accounting for damage, deterioration and modification based on
platform inspection records. Figure 2 presents the three-dimensional structural models
for Platforms A and B.
In this case study the non-linear pushover analyses were performed using the
USFOS program. The pushover analyses resulted in minimum RSR of 6.01 and 3.83
for Platforms A and B, respectively, when subjected to a wave from platforms
diagonal direction. It implies that Platforms A and B can resist 6.01 times and 3.83
times of the load imposed by the 100-year return period storm, respectively. Pushover
analyses also showed the collapse of both platforms was caused by a failure
mechanism formed within the jacket by brace buckling and yielding. No significant
non-linearities were observed in the piles or deck legs.

2010 Structures Congress 2010 ASCE

2776

(a)
(b)
Figure 2. Platform models: (a) Platform A; (b) Platform B
4.2

Reliability Analysis

A metocean study has been carried out using hind-casting techniques to


determine the frequency of hurricane events with different magnitudes for these two
platform locations. The extreme value distribution for significant wave height Hs
were derived from peaks-over-threshold extremal analysis. The significant wave
height Hs in storms was fitted using a Weibull distribution. Mean value and standard
deviation of Hs are given as
1

+ 1

H = +
2

(4)

H = + 1 2 + 1

(5)

where , and are Weibull distribution parameters: = 1.1687, = 2.0970 and


= 3.8624.
The annual probability of exceedance of significant wave height was found by
describing the frequency of occurrence of storms as a Poisson process. The annual
probability of exceedance of Hs is given as

Pe (H ) = 1 exp exp ((H )/)

(6)

where is a storm rate defined as = N storms / Thindcast , Nstorms is the number of storms in
the hindcase record, and Thindcast is the duration of the hindcast (years).

2010 Structures Congress 2010 ASCE

2777

The regression of maximum wave height Hm as a function of Hs gives


H m = 1.887 H s0.951

Multiple USFOS runs were performed for a number of wave heights to


generate wave loads. Coefficients C1, C2 and C3 were determined by simultaneously
fitting the curve in both regions above and below Hd using the regression method.
The regression of wave load to wave height yielded the parametric relationship of
wave load as a function of wave height. Relationships between the wave height and
wave/current load are plotted in Figure 3 for Platforms A and B.

Wave Load (100 x kN)

60
50

Platform A

40
30
20
10

Platform B

0
5

10

15

20

25

Annual Maximum Wave Height (m)

Figure 3. Wave load characteristics for Platforms A and B

The platform capacity R is assumed to have a lognormal distribution with a


coefficient of variation of 0.15 as in PMB report (1996). The coefficient C1 is
assumed as a lognormal random variable with a coefficient of variation of 0.1 as in
PMB report (1996). Given the large variability in the wave-in-deck forces, the
coefficient C3 is assumed as a lognormal random variable with a coefficient of
variation of 0.35 as recommended by API RP 2A. Table 2 summarizes the random
variables used in the reliability analysis and provides a description of the distribution
types and coefficients of variation (CoV).
Random Variable

Distribution

CoV

Lognormal

0.15 for jacket


0.20 for foundation

Weibull

per formula (4), (5)

Coefficient C1

Lognormal

0.10

Coefficient C3

Lognormal

0.35

Platform Capacity, R
Annual Significant Wave height, Hs

Table 2. Description of random variables

2010 Structures Congress 2010 ASCE

2778

For the location of these platforms in the Eastern Caribbean, the extreme
storm waves approach from an arc bounded by approximately 40 degrees and 90
degrees, measured clockwise from True North. For both Platforms A and B, the
extreme storm waves will dominantly approach from the platforms diagonal
direction. In this study only the annual failure probability for the diagonal wave
direction was considered, however for other regions, the annual failure probability
may be calculated for up to 8 wave directions. The annual probability of platform
failure has been calculated using FORM method. Platforms A and B have annual
probability of failure of 3.53210-5 and 1.43810-4, respectively.
4.3

Acceptance Criteria

To develop acceptance criteria, a series of reliability analyses have been


conducted to establish the relationship between the probability of platform failure and
the platform RSR by changing the platform RSR while keeping the same load pattern.
This process is to simulate different platform capacity in a platform fleet. The results
of probability of platform failure are presented in Figure 4. It is observed that using
this simple approach, there is good correlation between two curves. In theory,
different platforms have different configurations, and may be located at different
places with different water depth. They may also have different deck height (air gap)
and different foundation system. Therefore, the load pattern may vary across a
platform fleet. This simple approach provides an approximation on the probability of
failure for platforms with different RSR.

Annual Failure Probability, Pf

1.0E-01
Platform A as a basis
Platform B as a basis
1.0E-02

1.0E-03

1.0E-04

1.0E-05
0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

Reserve Strength Ratio, RSR

Figure 4. Probability failure versus RSR

For high-consequence manned existing platforms in the Gulf of Mexico, the


implied acceptable failure probability is 0.001 as shown in De (1995). The value of
0.001 is also the implied acceptable failure probability for high consequence existing
platforms offshore California. The study on the acceptance reliability criteria

10

2010 Structures Congress 2010 ASCE

2779

conducted by Stahl et al. (2000) also recommends an acceptable failure probability of


0.001 for an existing platform complex in the North Sea.
To be consistent with recommended or implied acceptance criteria in offshore
assessment practices in other parts of the world, the acceptable failure probability of
0.001 may be adopted for Eastern Caribbean offshore platforms. It is observed from
Figure 4 that utilizing the reliability parameters in Table 2, the target RSR value
required to meet the target probability of failure of 0.001 is 2.0 for the highconsequence category in the Eastern Caribbean offshore. Comparing to the target
RSR of 1.8 for typical North Sea platforms (Ronalds et al. (2003)) and 1.6 for typical
GoM platforms (API RP 2A (2005)), Eastern Caribbean offshore platforms have
relatively high target RSR values.
It is stressed that the above target RSR value is purely indicative, being for
particular locations in Eastern Caribbean offshore, water depth, structural
configuration and critical action. To derive a general acceptance criteria for Eastern
Caribbean offshore, more representative platforms should be studied and other
approaches for the development of target reliability should also be investigated.
4.4

Risk-Based Inspection

The annual probability of platform failure and the consequence of failure are
combined to determine the overall platform risk associated with hurricane hazard
based on a risk matrix in Figure 1. Both Platforms A and B are high consequence.
The likelihood of failure for Platform A is low while Platform B has medium
likelihood of failure. The risk evaluation indicates that Platform A is medium risk for
hurricane hazard and Platform B is high risk for the hurricane hazard. From Table 1,
the appropriate risk-based intervals for routine periodic inspections are 10 years and 5
years for Platforms A and B, respectively.
5

CONCLUSION
A comprehensive structural assessment approach has been presented for offshore
fixed platforms to ensure their continuous fitness-for-purpose. Non-linear pushover
analysis is used to calculate the platform capacity characterized in terms of the
platforms reserve strength ratio (RSR). Comparison of the platform RSR and the
target RSR specified in the acceptance criteria determine the fitness-for-purpose of
the platforms. The regionally specific acceptance criteria can be developed based on
structural reliability methodology. The assessment results can also be used to develop
a effective long term risk-based underwater inspection strategy. A case study for the
platforms located offshore Eastern Caribbean has illustrated the effectiveness of the
developed structural assessment approach.

REFERENCES

American Petroleum Institute (API) (2005). Recommended Practice for Planning


Designing and Constructing Fixed Offshore Platforms - Working Stress
Design. API RP 2A-WSD 21st Edition, Errata and Supplement 2.

11

2010 Structures Congress 2010 ASCE

2780

Botelho, D. L. R., and Petrauskas, C. (1994). A Detailed Study on the Failure


Probability of ST130 A Platform During the Passage of Hurricane
Andrew. OTC7472, 26th Offshore Technology Conference, Houston, Texas.
Gebara, J., Dolan, D., Pawsey, S., Jeanjean, P., and Dahl-Stamnes, K. H. (2000).
Assessment of Offshore Platforms Under Subsidence-Part I: Analysis and
Results. Journal of Offshore Mechanics and Arctic Engineering, Vol. 122, pp
260-266.
Jha, A. K., Kiciman, O. K., Gebara, J. M., Stahl, B., and Dahl-Stamnes, K. H. (2000).
Assessment of Offshore Platforms Under Subsidence-Part II: Analysis and
Results. Journal of Offshore Mechanics and Arctic Engineering, Vol. 122, pp
267-273.
Heredia-Zavoni, E., Campos, D., and Ramirez, G. (2004). Reliability Based
Assessment of Deck Elevations for Offshore Jacket Platform. Journal of
Offshore Mechanics and Arctic Engineering, Vol. 126, pp 331-336.
Dier, A. F., Morandi, A. C., Smith, D., Birkinshaw, M., and Dixon A. (2001). A
Comparison of jacket and jack-up structural reliability. Marine Structures,
Vol. 14, pp 507-521.
Ronalds, B. F., Anthony, N. R., Tuty, S., and Fakas, E. (2003). Structural reliability
of monopods under storm overload. Journal of Offshore Mechanics and
Arctic Engineering, Vol. 125, pp 114-118.
Moan, T. (2000). Recent research and development relating to platform
requalification. Journal of Offshore Mechanics and Arctic Engineering, Vol.
122, pp 20-32.
Onoufriou, T., and Forbes, V. J. (2001). Developments in structural system
reliability assessments of fixed steel offshore platforms. Reliability
Engineering and System Safety, Vol. 71, pp 189-199.
American Petroleum Institute (API). Recommended Practice for the Structural
Integrity Management of Fixed Offshore Structures. API RP 2SIM (Draft).
Stahl, B., Aune, S., Gebara, J. M., and Cornell, C. A. (2000). Acceptance Criteria for
Offshore Platforms. Journal of Offshore Mechanics and Arctic Engineering,
Vol. 122, PP. 153-156.
UK HSE (2002). Marine Risk Assessment. OTR 2001-063.
SINTEF (2008). USFOS Non-Linear Program. Version 8.0, Norway.
PMB Engineering, Inc. (1996). Hurricane Andrew Effects on Offshore Platforms,
Phase II Joint Industry Project. Final Report, San Francisco, CA.
De, R. S. (1995). Risk Analysis Methodology for Developing Design and
Assessment Criteria for Fixed Offshore Structures. OTC Paper No. 7755,
Proc. Offshore Technology Conference, pp. 543-556.

12

You might also like