ABS TECHNICAL PAPERS 2005

Similar documents
Life Extension of Mobile Offshore Units

RULES FOR CLASSIFICATION Ships. Part 3 Hull Chapter 9 Fatigue. Edition October 2015 DNV GL AS

Development of TEU Type Mega Container Carrier

Presented to the Israel Annual Conference on Aerospace Sciences, 2009 RISK-ANALYSIS A SUPPLEMENT TO DAMAGE-TOLERANCE ANALYSIS

Guide for Assessing Hull-Girder Residual Strength for Tankers. July 1995

Technical Standards and Legislation: Risk Based Inspection. Presenter: Pierre Swart

Special Considerations for Structural design and Fabrication for. tankers or similar vessels with Large Size (150m or more in length) in.

Risk-Based Condition Assessment and Maintenance Engineering for Aging Aircraft Structure Components

Abstract. 1 Introduction

MSC Guidelines for Review of Gas Carrier/Barge Structures

Modelling of Extreme Waves Related to Stability Research

Failure Data Analysis for Aircraft Maintenance Planning

ICE LOADS MONITORING SYSTEMS

MAE 322 Machine Design Lecture 5 Fatigue - 2. Dr. Hodge Jenkins Mercer University

IACS URS11 defines the dimensioning wave load for ship design, but what does it mean from a statistical point of view?

Abstract. 1 Introduction

Learn more at

Justification of Risk Reduction through In-Service Inspection / REDUCE

Vibration and Pulsation Analysis and Solutions

Enbridge Pipelines Inc. PIPELINE INTEGRITY AXIAL CRACK THREAT ASSESSMENT

Discussion on the Selection of the Recommended Fish Passage Design Discharge

PASSENGER SHIPS Guidelines for preparation of Hull Structural Surveys

Combined Wave and Wind Fatigue Damage for Offshore Structures

The Benefits Of Composite Materials In Deepwater Riser Applications. 26 th March 2015 Hassan Saleh Senior Engineer 2H Offshore Engineering Ltd

ASPIRE for Integrity Management Support for Upstream Assets. Payam Jamshidi, TWI Ltd Sebastian Hartmann, Innospection Ltd

Part 1: General principles

ESTIMATION OF THE DESIGN WIND SPEED BASED ON

CONVECTION SECTION FAILURE ANALYSIS AND FITNESS-FOR-SERVICE ASSESSMENT

NKS PODRIS project. Importance of inspection reliability assumptions on piping failure probability estimates

ASSESSING THE RELIABILITY OF FAIL-SAFE STRUCTURES INTRODUCTION

RULES PUBLICATION NO. 52/P UNDERWATER INSPECTION OF MOBILE OFFSHORE DRILLING UNITS IN LIEU OF DRYDOCKING

POWER Quantifying Correction Curve Uncertainty Through Empirical Methods

An approach to account ESP head degradation in gassy well for ESP frequency optimization

Residual Strength of Damaged Ship Hull

Introduction. DEEPWATER DRILLING RISER INTEGRITY - FATIGUE, WEAR, INSPECTION AND MONITORING by Dr Hugh Howells and Dave Walters 2H Offshore Inc

OTC MS. Free Span Rectification by Pipeline Lowering (PL) Method N. I. Thusyanthan, K. Sivanesan & G. Murphy

ITTC Recommended Procedures Testing and Extrapolation Methods Loads and Responses, Seakeeping Experiments on Rarely Occurring Events

Engineering Practice on Ice Propeller Strength Assessment Based on IACS Polar Ice Rule URI3

Control Strategies for operation of pitch regulated turbines above cut-out wind speeds

TG GUIDELINES CONCERNING CALIBRATION INTERVALS AND RECALIBRATION

OMAE COST EFFECTIVENESS OF HULL GIRDER SAFETY

Fatigue Resistant Sockets for Dynamic Loads PFEIFER Cable Structures. Cable structures technology Info Nr PFEIFER SEIL- UND HEBETECHNIK GMBH

The Challenge of Wave Scouring Design for the Confederation Bridge

for Naval Aircraft Operations

Inspection Credit for PWSCC Mitigation via Peening Surface Stress Improvement

Determination of the Design Load for Structural Safety Assessment against Gas Explosion in Offshore Topside

Application of Simulation Technology to Mitsubishi Air Lubrication System

Fitness for Service Assessment of Ageing Pressure Vessel Experiencing External Corrosion: A Case Study

Kiefner & Associates, Inc.

Risks Associated with Caissons on Ageing Offshore Facilities

Next Generation Quartz Pressure Gauges

SPE Copyright 2001, Society of Petroleum Engineers Inc.

INTERNAL AXIAL CORROSION IN OFFSHORE PIPELINES: INSPECTION AND ASSESSMENT. By: Dr. Christoph Jäger, Abdullahi Atto, NDT Global, Germany

Chapter 5: Methods and Philosophy of Statistical Process Control

ShipRight Design and construction. Procedure for Ship Units July 2014

Proceedings of the ASME th International Conference on Ocean, Offshore and Arctic Engineering OMAE2011

Impact Fatigue on Suction Valve Reed: New Experimental Approach

2 FUSION FITTINGS FOR USE WITH POLYETHYLENE PRESSURE PIPES DESIGN FOR DYNAMIC STRESSES

Influence of wave steepness on extreme ship hull vertical wave bending moments

RIGID RISERS FOR TANKER FPSOs

Statistics of still water bending moment of damaged ships

Assessment Summary Report Gulf of Mexico Red Snapper SEDAR 7

Offshore platforms survivability to underwater explosions: part I

J. J. Daly Metal Improvement Company Paramus New Jersey USA

INTEGRITY MANAGEMENT OF STRESS CORROSION CRACKING IN GAS PIPELINE HIGH CONSEQUENCE AREAS

EFFECTS OF SIDEWALL OPENINGS ON THE WIND LOADS ON PIPE-FRAMED GREENHOUSES

FATIGUE ASSESSMENT OF OFFSHORE STRUCTURES

Monitoring or Inspection

WAVE IMPACTS DUE TO STEEP FRONTED WAVES

(DD/MMM/YYYY): 10/01/2013 IP

E. Agu, M. Kasperski Ruhr-University Bochum Department of Civil and Environmental Engineering Sciences

Liquefied gas cargo tanks and process pressure vessels

EXPERIMENTAL STUDY ON THE HYDRODYNAMIC BEHAVIORS OF TWO CONCENTRIC CYLINDERS

On The Relationship Between Missing Shoulder Ballast and Development of Track Geometry Defects

Wire ropes condition monitoring: conception and embodiment

Marine Construction & Welding Prof. Dr. N. R. Mandal Department of Ocean Engineering & Naval Architecture Indian Institute of Technology, Kharagpur

Risk Based Underwater Inspection (RBUI) For Existing Fixed Platforms In Indonesia

Designing an ICCP System for the Hull of an Arctic Ice-Breaking Vessel

The Application Of Computer Modeling To Improve The Integrity Of Ballast Tanks

Development of Accidental Collapse Limit State Criteria for Offshore Structures

Damage Assessment Following Accidents - Official Discussion

A Study on Roll Damping of Bilge Keels for New Non-Ballast Ship with Rounder Cross Section

A FINITE ELEMENT PROCEDURE TO MODEL THE EFFECT OF HYDROSTATIC TESTING ON SUBSEQUENT FATIGUE CRACK GROWTH

Feasibility of Steel Lazy Wave Risers in the North Sea

PRAGMATIC ASSESSMENT OF EXPLOSION RISKS TO THE CONTROL ROOM BUILDING OF A VINYL CHLORIDE PLANT

EVALUATING CRITERIA FOR DP VESSELS

Effects of directionality on wind load and response predictions

TLP Minimum tendon tension design and tendon down-stroke investigation

Appendix E Cat Island Borrow Area Analysis

RESOLUTION MEPC.64(36) adopted on 4 November 1994 GUIDELINES FOR APPROVAL OF ALTERNATIVE STRUCTURAL OR OPERATIONAL ARRANGEMENTS AS CALLED FOR IN

Tightening Evaluation of New 400A Size Metal Gasket

Sea-going vessel versus wind turbine

PREDICTING THE ABILITY OF SURVIVAL AFTER DAMAGE IN TANKERS. José Poblet Martínez - SENER, (Spain) José Javier Díaz Yraola - SENER, (Spain)

HIGH FLOW PROTECTION FOR VARIABLE SPEED PUMPS

NASA AEROSPACE PRESSURE VESSEL SAFETY STANDARD

RULES FOR CLASSIFICATION Ships. Part 3 Hull Chapter 5 Hull girder strength. Edition October 2015 DNV GL AS

EXPERIMENTAL RESULTS OF GUIDED WAVE TRAVEL TIME TOMOGRAPHY

SPE The paper gives a brief description and the experience gained with WRIPS applied to water injection wells. The main

Author s Name Name of the Paper Session. Positioning Committee. Marine Technology Society. DYNAMIC POSITIONING CONFERENCE September 18-19, 2001

Transformer fault diagnosis using Dissolved Gas Analysis technology and Bayesian networks

Transcription:

OTC 17535 Structural Reliability Applications in Risk-Based Inspection Plans and Their Sensitivities to Different Environmental Conditions A. P. Ku and R. E. Spong, Energo Engineering; C. Serratella, ABS Consulting, Inc.; X. S. Wu, ExxonMobil Upstream Research Company; R. Basu and G. Wang, American Bureau of Shipping. Copyright 2005, Offshore Technology Conference This paper was prepared for presentation at the 2005 Offshore Technology Conference held in Houston, TX, U.S.A., 2 5 May 2005. This paper was selected for presentation by an OTC Program Committee following review of information contained in a proposal 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 a proposal of not more than 300 words; illustrations may not be copied. The proposal 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. Abstract Structural system inspection practices for floating offshore facilities, such as FPSO s and FSO s, have for the most part followed the practice employed and developed via trading ships/maritime practice. Often, the influence of a particular repeated service/cargo loading regime, cargo corrosivity, or site environmental influence can play a major role for a particular asset. More specifically, there has been only cursory attempts made to first understand and then, where possible, take advantage of site-specific conditions and loading in the development of in-service inspection plans. In recent years there has been significant interest by the marine and offshore industry to apply structural reliability techniques to risk based inspection planning for marine vessels and floating production installations. In this regard, structural reliability based methods can assist in providing a framework for quantifying sitespecific loading and degradation mechanisms (such as fatigue and corrosion), through a systematic consideration of the probabilistic uncertainty in each degradation mechanism. By applying structural reliability analysis and risk assessment techniques to inspection planning, the operator is given a tool by which he can justify the allocation of resources to those structural components with a higher risk profile, and at the same time potentially relax inspection activities for low risk components to optimize and target inspection efforts. In a companion paper [1], available structural reliability methods developed to date were summarized, and then applied to determine the inspection intervals based on site specific loading as applied to strength considerations of the hull girder as well as to stiffened and un-stiffened plate panels. By tracing the timevarying reliability index for these structural components, the risk-based inspection intervals can be determined. This methodology has recently been implemented to provide the foundation in a risk-based inspection (RBI) plan for a floating production unit offshore West Africa. The current paper will further consider the sensitivities resulting from the following two conditions for the strength reliability: (1) environmental load for different regions of the world, and (2) corrosion rate corresponds to offshore environment as well as storage conditions. Another two conditions are considered for fatigue reliability: (3) assumed initial crack size for a connection, and (4) crack growth parameter for either the air environment or marine environment that a connection may be subjected to. The sensitivity studies presented in this paper provide a quick reference to understand the RBI plans benefits via work scope optimization and cost reductions for floating production units operating in various regions of the world. Introduction Hull Girder Reliability Analysis (First-Yield & Ultimate) Sea Keeping Analysis and Generation of Load Cases Global FEM Analysis under Current Scantling Strength Reliability Assessment Selection of Structural Components to Perform Structural Reliability Stiffened/Un-Stiffened Panel Reliability Analysis Information Gathering and Quantification of Degradation Mechanisms Hazard Identification Exercise (HAZID) to Identify Higher Risk Components S-N Curve Based Fatigue Reliability Analysis Determination of Target Reliability Level Fatigue Reliability Assessment Risk Based Inspection Extent and Schedule Figure 1 - A risk-based inspection planning reliability analysis flow chart [1] Fracture Mechanics Based Fatigue Reliability Analysis Correlation Study on Similar Connections Structural Reliability Applications in Risk-Based Inspection Plans & Sensitivities to Different Environment Conditions 17

Details of a structural reliability methodology applicable to risk based inspections on a ship-shaped floating production vessel have been published in a companion paper [1]. From that paper, a flow chart of this methodology is shown in Figure 1. This methodology starts with information gathering, such as obtaining the vessel gauging data, vessel past history, current service condition, etc. For the example vessel considered in [1], direct seakeeping and global finite element analysis were conducted to identify representative high-stressed structural components of all types for further structural reliability analyses. The model plot of this global finite element analysis is shown in Figure 2. the current application can be put into perspective. The highest consequence (Consequence I) implies the total loss of vessel plus the loss of production for a time period of approximately two years. Consequence II is one order of magnitude less (i.e., represents significant damage) than Consequence I in terms of financial loss, and similarly for Consequence III. The target reliabilities are also compared to the published Ship Structure Committee recommended values for tankers [2], which is presented in Table 1. Note that in Table 1, Consequence I is further divided into I-a and I-b, in which Consequence I-a applies to reliability analysis if the hull girder first-yield capacity is considered. Consequence I-b applies to hull girder ultimate capacity analysis. Since first-yield capacity typically overestimates the true capacity of the hull girder, a higher reliability level is required. The Consequence I as shown in Figure 3 is for Consequence I-b for the ultimate capacity, which represents the true capacity of the hull girder. 1.0E+00 1.0E-01 Example Vessel RBI Consequence III Unacceptable Region 1.0E-02 MODUs Annual Pf 1.0E-03 Foundations Fixed Drilling Rigs Example Vessel RBI Consequence I Figure 2 - Global FEM model of an example FPSO [1] The strength reliability analyses were carried out both on the global strength level and local strength level. For global strength analysis, several frames along the length of the vessel were selected and their firstyield strength as well as the ultimate strength was analyzed. Corrosion rates and the wave-induced bending moment were considered as random variables, among others, for the global strength analysis. For local strength analysis, a representative number of plate panels were selected with the objective to envelope the load and expected corrosion conditions along the entire length of the vessel. Similar to the hull girder strength analysis, the primary stress component due to wave-induced motion, together with corrosion rates applicable to the plate and stiffeners, were considered as the main random variables. The strength reliabilities for both the hull girder and plate panels were calculated on an annual basis and compared to a chosen target value to determine if the reliability values (reliability index) are acceptable. The target values were determined from a risk diagram as shown in Figure 3. Figure 3 combines the consequence and likelihood information on one chart, which also conveniently overlay the experiences from other industries such that 1.0E-04 1.0E-05 Acceptable Region Buildings Commercial 1.0E-06 Aviation 0.01 0.1 1 10 100 1000 10000 Dams Example Vessel RBI Consequence II Consequences of Failure (MM$) Fixed Platforms Naval Combatants Figure 3 - Risk diagram to determine acceptable annual probability for various financial consequences [1] Table 1 - Lifetime target reliability comparison between example vessel and Ship Structure Committee report [2] recommended values Strength Assessment Fatigue Assessment Consequence Note: Example Vessel Ship Structure Committee Example Vessel Ship Structure Committee I-a 4.50 4.50 N/A N/A I-b 4.00 4.00 N/A N/A II 3.20 3.50 3.20 3.00 III 2.50 3.00 2.50 2.50 IV 1.50 N/A 1.50 2.00 I-a refers to hull girder first-yield reliability I-b refers to hull girder ultimate reliability In summary, the consequence levels (i.e., I-a, I-b, II, III and IV) are associated with the lifetime target reliabilities such that specific strcutural damage (i.e., Marginally Acceptable 18 Structural Reliability Applications in Risk-Based Inspection Plans & Sensitivities to Different Environment Conditions

fatigue or strength) should not drop below the assigned target probability. Part I: Strength Reliability Sensitivity This section focuses on the strength reliability of an example plate panel in a ship-shaped production vessel. The plate panel considered is a longitudinal bulkhead (LBH) plate located toward the bottom plating. This LBH plate is shown in Figure 4, which is subject to relatively high stress from hull girder bending moments. There are other relatively high stressed areas shown in Figure 4, for example toward the aft end of the LBH. However, stresses in this region are dominated primarily from the geometry change and less so from the environmental loads. The reliability limit state function of this LBH plate can be written as follows: g = B R B S (1) r s where R and S represent the probabilistic resistance and load variables. BR and BS are the model uncertainty factors for the resistance and load calculations, respectively. The details on the probabilistic distributions for BR and BS can be found in [1]. The resistance variable, R, is a set of plate buckling equations, the details of these equations can be found in Reference [3]. In these equations the main variables are the geometry of the plate panel, the number of stiffeners within the plate panel, and thickness of the plate and stiffeners. The corrosion rates have a direct impact on the plate and stiffener thickness and whose sensitivity will be studied in the following reliability analysis. The geometry of the plate panel is of the typical size found between two web frames encountered in a traditional VLCC tanker. The load variable, S, consists of the wave-induced and still water induced bending moments. For simplicity, the still water-induced bending moment is treated as a fixed value among all sensitivity cases studied. The differences in load variable S are then primarily from wave environmental load and will be distinguished among four major offshore production regions of the world: West Africa, Brazil, Gulf of Mexico and North Sea in the following sensitivity study. Analyzed Panel Figure 4 - A longitudinal bulkhead plate used for strength reliability sensitivity analysis Environmental Load Reference [4] performed sea-keeping calculations to estimate the maximum vertical wave bending moment (VWBM) acting on the mid-section hull girder of a tanker-shaped floating vessel. The ratios between the maximum VWBM for four major offshore production regions of the world are shown in Table 2. Note that in Table 2 the ratios are normalized with respect to the West African (lowest) value. Table 2 Ratio of Vertical Wave Bending Moment of a 315m Tanker Shaped Vessel West Africa 1.00 Brazil 1.27 GOM 1.40 North Sea 2.04 The primary stress component (longitudinal stress) on the LBH plate panel is assumed to be in direct proportion to the VWBM. In the reliability analysis the longitudinal stress is modeled as a Type-I extreme value distribution [1]. Also the distribution parameters of this distribution were calculated from direct sea-keeping analysis with West African wave conditions [1]. In the current reliability sensitivity study, the longitudinal stress distributions have mean values scaled by the ratios of Table 2, which the coefficient of variations are maintained the same for the four offshore regions considered. Corrosion Rate In addition to the four environmental cases, the influence of three representative corrosion rates (i.e., high, medium and low) are also evaluated in the study. The rate distributions are assumed to be represented by the Weibull distribution function. Table 3 shows the three assumed rates statistics (i.e., mean and coefficient of variation (c.o.v.)). Structural Reliability Applications in Risk-Based Inspection Plans & Sensitivities to Different Environment Conditions 19

Table 3 Weibull Probabilistic Corrosion Rates Used for LBH Structural Elements Low Corrosion Medium Corrosion High Corrosion Mean (mm/year) c.o.v. 0.08 50% 0.16 50% 0.30 50% The mean rates shown in Table 3 are assumed for all of the structural elements (i.e., LBH plating, longitudinal webs and flanges). Furthermore the mean rate represents total material wastage for both sides of the structural element. This assumes that all elements, regardless of there orientation will have the same rate of material loss, and that both sides of the LBH plating is exposed to the same service conditions (i.e., crude oil storage tanks on both sides of LBH). The three rates in Table 3 represent uniform corrosion for an uncoated surface exposed to crude oil. The low and medium corrosion are from the Guidance Manual for Tanker Structures [8] suggested corrosion rate range for LBH plating and longitudinal webs. The low corrosion represents the lower range rate, and medium corrosion represents upper range rate. The high corrosion is approximately twice that of the upper rates suggested in Reference 8, and it is intended to represent a severe general corrosion case. For all rates, the c.o.v. is assumed to be 50%. Corresponding statistical values for oil tankers can be found in, e.g., [5]. In [5], and other accompanying references in [5], many thousands of gauging data from a wide range of tankers were analyzed with statistical distribution parameters calculated. The c.o.v. from those analyses is in the range of 100% for a wide range of structural components. However, the data was gathered from a wide spectrum of tankers, potentially with very different tank service conditions and integrity management philosophies. In FPSO service for a single vessel, we expect the coefficient of variation to be smaller than the values given in [5]. Combining the four cases for environmental load conditions and three corrosion rate cases, there are a total of 12 cases considered in this strength reliability sensitivity study. These 12 cases are listed in Table 4 for easy reference. Table 4 Case Number for Different Combinations of Environmental Load and Corrosion Rate Low Medium High Corrosion Corrosion Corrosion West Africa 1 2 3 Brazil 4 5 6 GOM 7 8 9 North Sea 10 11 12 Strength Reliability Variations for Sensitivity Cases Considered Figure 5 shows the strength reliability sensitivities for Cases 2, 5, 8 and 11 in which the corrosion rates were set to the medium level, with the environmental loads varied from West Africa to North Sea wave conditions. Results from Figure 5 indicate that the example LBH plate panel has a sufficient reliability for almost the entire duration of design life (20 years) for West Africa, Brazil and GOM. In terms of scheduling of the risk based inspection interval, this plate panel and all other panels with smaller stress will not dominate the inspection interval. Also, the strength reliability is found to be not sensitive among these three environmental conditions. The strength reliability for the North Sea environmental wave condition is found to be substantially lower than the other three environmental regions. A plate panel which might not dominate the RBI schedule consideration in the first three regions can become a dominant issue when the vessel is located in the North Sea. Reliability Index (beta) 5 4.5 4 3.5 3 Annual Strength Reliability of a LBH Panel (Medium Corrosion Rate) Consequence III Target 2.5 0 5 10 15 20 Time (Year) Case 2, West A. Case 5, Brazil Case 8, GOM Case 11, North Sea Figure 5 Annual Strength Reliability for Four Environmental Loads with Medium Corrosion Rate It needs to be pointed out that the target reliability level shown in Figure 5 is the annual target reliability, which is 3.4 for the Consequence III category. In Table 1 where the target reliabilities for various consequences are listed, the lifetime target reliability values are used for the purpose of convenient comparison to [2]. Figure 6 shows the strength reliability sensitivities for Cases 7, 8 and 9 in which only the GOM environmental load is considered with the corrosion rates varied from low to high levels. Results from Figure 6 show that the corrosion rates have major impacts on the strength reliability. For example, Case 7 (low corrosion rate) passes through the acceptable level at roughly year 50 (extrapolated beyond shown in Figure 6), while Case 8 (medium rate) passes through at year 18 and Case 9 (high rate) passes through at year 10. Also as observed in Figure 6, the rate of strength reliability deterioration is very different for the three corrosion rate cases with the high corrosion rate case 20 Structural Reliability Applications in Risk-Based Inspection Plans & Sensitivities to Different Environment Conditions

falling much more precipitately than the other two cases. Reliability Index (beta) 4.5 4.3 4.1 3.9 3.7 3.5 3.3 3.1 2.9 2.7 Annual Strength Reliability of a LBH Panel (GOM Wave Environment) Consequence III Target 2.5 0 5 10 15 20 Time (Year) Case 7, Low Rate Case 8, Med Rate Case 9, High Rate Y (a) is the geometry function of the crack shape, C i, m are the crack growth parameters of the material, ν 2 is the stress range annual frequency (cycles/year), T is the time under consideration (year), ε Y is the model uncertainty for geometry function, ε S 2 is the model uncertainty for stress range calculation, A, B are the Weibull stress range parameters. Figure 6 Annual Strength Reliability for Three Corrosion Rates with GOM Environmental Load It is found from the combined information given in Figures 5 and 6 that the RBI schedule has more sensitivity with respect to corrosion rate than to environmental load conditions. A vessel located in the North Sea can be kept with a high strength reliability level throughout the service life with lower corrosion rates, even though the environmental load is much more demanding compared to the other regions of the world. Note that the effects of low corrosion rate uncertainty can be achieved in two independent ways: by lowering the expected (mean) corrosion rate or by lowering the spread (c.o.v.) of the corrosion uncertainty. For example, the same reliability can be achieved between a corrosion rate random variable with a mean value of 0.08 mm/year with a 50% c.o.v. and another corrosion rate with mean value of 0.3 mm/year but a smaller c.o.v. (e.g. 20%). This implies that even for vessels with high expected rates, due to anticipated tank service conditions (e.g., high cargo temperatures, infrequent tank usage, etc.), high reliability can still be achieved by implementing methods to closely monitor the corrosion rates observed in order to reduce the uncertainty associated with this high rate. Part II: Fatigue Reliability Sensitivity In this section, the fatigue reliability sensitivity with respect to different initial crack sizes and crack growth parameters are examined. Figure 7 illustrates surface cracks propagated from a weld toe. Fatigue reliabilities were calculated using fracture mechanics formulation, with the failure criterion defined as a through-thickness crack. The development of fracture mechanics based fatigue reliability equations can be found in, e.g., [1], [6]. The limit state function reads: M a di da = m m m t) Ci ν 2T ε S A Γ(1 + ) m (2) B ( ε YY ( a) πa ) i ( 2 where a oi a di a oi is the initial crack depth, is the final crack depth at fatigue failure, Figure 7 Illustration of Cracks at Weld Toe Cases Studied Two different sizes of initial crack depth are considered with a constant aspect ratio 1/20: Small Initial Crack crack depth less than 0.3mm with 95% probability. Large Initial Crack crack depth less than 1.5mm with 95% probability. In addition to different initial crack sizes, two types of crack growth parameters are considered, with the first type corresponds to the air environment (used within voids or non-water-wetted surfaces), while the second type corresponds to the marine environment with cathodic protection at -850 mv [7]. Table 5 summarizes the cases considered. Table 5 Sensitivity Cases for Fatigue Reliability Variations Small Initial Large Initial Crack Crack Air Environment Case 1 Case 2 Marine Environment Case 3 Case 4 Figure 8 shows the time-variant fatigue reliabilities of the example connection for the four sensitivity cases. The results show Case 3 drops more from Case 1 than Case 2 drops from Case 1. In other words, the crack growth parameters for different environment appear to have a larger impact on the fatigue reliability than the size of the initial crack. Structural Reliability Applications in Risk-Based Inspection Plans & Sensitivities to Different Environment Conditions 21

Fatigue Reliability (beta) 6 5 4 3 2 1 0 Consequence III Target 0 5 10 15 20 Time (Year) Figure 8 Lifetime Fatigue Reliability Variations for Sensitivity Cases Studied Case 1 Case 2 Case 3 Case 4 The fatigue reliability updating for all four sensitivity cases were carried out using the formulations as presented above and assuming the inspection has an accuracy equivalent to the MPI inspection. Furthermore, it is assumed that no cracks were found in all inspections. Figure 10 shows the results for Case 3, in which inspections were carried out immediately after the time when the fatigue reliability is predicted to fall below the acceptable level. In the Case 3 example, inspections need to be carried out at the 2 nd, 4 th, 7 th, 10 th, 14 th and the 18 th year. It is seen from Figure 10 that once inspections are carried out without finding any cracks, the fatigue reliability index increases dramatically right after the inspection, and then decreases as time elapses. Sensitivity on Inspection Updating The probability of fatigue failure at year t, with inspection performed at year t ( t t ) is inspect inspect calculated by the following conditional probability equations: P [ M ( t) 0 I, ( t ) > 0] (3) i < no i inspect for the no cracks found scenario, and P [ M ( t) 0 I, ( t ) = 0] (4) i < yes i inspect for the crack found and measured scenario. The probability of detection used in this study was adopted from the information contained in [6]. These detection probabilities for the close visual and the MPI inspection are shown in Figure 9. With reference to this figure, the probability of detection (POD) can be represented by the following Exponential distribution: F ( x) = 1 exp( λ ( x τ )) (5) The distribution parameters for the close visual and the MPI inspections can be found in [1]. Figure 10 Effects of Inspection Updating for Case 3 Crack The time instances when inspections need to be carried out are summarized in Table 6 for all sensitivity cases. In this Table, the years necessary for inspections as predicted from fatigue reliability analysis are marked. It is observed from Table 6 that once inspections are carried out without finding cracks, the effects of assumed initial crack sizes are greatly diminished. For example, after the 1 st inspection for Cases 1 and 2, although occurring at different time instances, the next inspection cycles follow a seven to eight years frequency for both cases. A similar phenomenon applies to the comparison between Cases 3 and 4. Case 1 Case 2 Case 3 Case 4 Table 6 Example Inspection Interval for Sensitivity Case Studies Year for Inspection 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Figure 9 - Probability of Detection for the Close Visual Inspection and MPI Inspection [6] Table 6 also indicates that for connections subjected to the Marine Environment, more frequent inspections are required compared to connections exposed to the Air Environment. The increased frequency for the Marine Environment connections is approximately two 22 Structural Reliability Applications in Risk-Based Inspection Plans & Sensitivities to Different Environment Conditions

to three times more compared to the Air Environment connections. Although from Table 6 it is predicted that Cases 3 and 4 need to be inspected on a three to four year cycle, it is important to point out that inspections can be rotated among a group of similar connections such that between every three to four years the inspections do not have to be done on the same set of connections. Correlations between a large sample of connections can be utilized such that only a percentage of total connections are inspected during the first inspection. If the inspection shows favorable results then by way of statistical correlation the remaining number of connections will have sufficient fatigue reliability. For the next inspection three to four years later, a second group of similar connections are then inspected. Based on these considerations mentioned above, the more unfavorable Marine Environment does not necessarily result in a significant increase in inspection cost. By intelligently grouping the connections and rotating the inspection schedule among them, high fatigue reliability for these connections can still be maintained. Conclusion The conclusions found from this study can be summarized as follows: Corrosion rates have a higher impact on strength reliability than environmental load conditions when considering the strength driven degradation (corrosion causing either yielding or buckling). High probabilistic corrosion rates show a very unfavorable trend to precipitate the reliability curve while the harsh environmental load shows a more favorable trend of maintaining the same slope as other more benign environmental conditions. High strength reliability for the harsh environmental load conditions can be achieved by reducing corrosion rates, which in turn can be achieved by - Implementing corrosion control measure to reduce the expected corrosion rates, and/or - Closely monitoring corrosion rate to reduce their uncertainty. The crack growth parameters between the Air Environment and Marine Environment (with the same initial crack size) have a larger impact on different initial crack sizes (with the same crack growth parameters). From fatigue reliability updating analysis, assumed initial crack sizes have little effect on subsequent inspection schedule once the first inspection is conducted. The subsequent inspection schedule (after the 1 st inspection) for the Maine Environment exposed connections needs to be two to three times more frequent than the Air Environment connections. When statistical correlations among like connections (addressed in a companion paper, [1]) are considered, rotation of inspections on subgroups for each scheduled inspections can reduce inspection cost while still maintaining necessary fatigue reliability. Acknowledgements The authors wish to thank their respective companies for the opportunity to publish this paper. References 1. Ku, A., Serratella, C., Spong, R., Basu, R., Wang, G., Angevine, D., Structural Reliability Applications in Developing Risk-Based Inspection Plans for a Floating Production Installation, Proceedings of the Offshore Mechanics and Arctic Engineering Conference, OMAE2004-51119, Vancouver, 2004. 2. Ship Structure Committee, Probability Based Ship Design: Implementation of Design Guidelines, SSC-392, 1996. 3. American Bureau of Shipping, Rules for Building and Classing Steel Vessels, 2004. 4. Xu, L., Nick, C. and Shin, Y., Implementation of Hindcast Global Wave Data for Design and Analysis of Marine Structures, Deep Offshore Technology, 2000. 5. Wang, G. and Spong, R., Experience Based Data for FPSO s Structural Design, Proceedings of the Offshore Technology Conference, OTC 15068, 2003. 6. Ship Structure Committee, Ship Maintenance Project: Fatigue Damage Evaluation, SSC-386-I, 1995. 7. BSI, Guide on Methods for Assessing the Acceptability of Flaws in Metallic Structures, BS 7910:1999. 8. TSCF (1997). Guidance Manual for Tanker Structures, Tanker Structure Co-operative Forum, Witherby & Co. Ltd. Structural Reliability Applications in Risk-Based Inspection Plans & Sensitivities to Different Environment Conditions 23