You are on page 1of 8

13th Congress of Intl. Maritime Assoc. of Mediterranean IMAM 2009, stanbul, Turkey, 12-15 Oct.

2009

Deepwater marine riser systems lifetime and fracture integrity prediction and assurance
A.M. Suliman
(PDI - Nipetco, Cairo, Egypt)

M.M. El-Gammal, Y.A. Abdel-Nasser & A.M. Rashwan


(University of Alexandria, Alexandria, Egypt) ABSTRACT: Few years ago, it has been noted that the the oil wells in the vicinity of the shore are getting dry and dry every day. So, it is vital now to realize the impact of searching for new wells in the very deepwater areas. Deepwater needs dictate that tighter tolerances on base materials and increased tolerances of accepted flaws being deployed thus deviating from the recommended codes and the more leaning towards rationally as well as the based optimized design approaches and the critical tolerable flaw studies without compromising the safety and integrity of the structure should be required. Many codes have come to recognize this approach and placed technical statements and reviewed guidelines. The intent of the present paper is to highlight and cover technical and theoritical tools of predictions of lifetime estimations based on brittle fracture inititiations and collapse countermeasures in accordance with codes guidelines on flaw acceptance which in turn are based on engineering criticality studies. Practical case studies have been included to technically demonstrate and prove the simplicity as well as the adaptaion of the application of the proposed approaches and the accuracy of the predicted estimates. developed and used. Deepwater offshore oil fields are a typical example of such fields due to its challenging nature which is progressively being overcome by emerging enabling technologies. This is clearly as shown in Figure 1. Typical challenges associated with deepwater offshore explorations and developments are feasibly seen in the increased pressures, decreased temperatures, harder floorbeds, higher current motions and more seabed magnitudes of waves. In addition to the common applications of the deepwater marine risers are also challenged by increased top tension, reduced buoyancy materials efficiency, increased vibrations and increased internal fluids pressures. The need for more stronger, more viable reduction exploitation times and the lesser maintenance periods have all been devoted to study structural integrity of offshore structures to emphasize both the verification of functions as well as the lifetime validity. The avialability of cost effective computer technologies has led to increasing application of risk and fitness for performance and fitness for service based design

1 INTRODUCTION The world demand for energy is ever increasing and the adequacy of energy supply is sharply challenged, particularly the hydrocarbon based fuel resources. The international reserve of mineral hydrocarbon resources of recoverable amount of the mineral resources available to disposition is heavily degraded and recently deteriorated. This is the anticipated results that account for both technical and economical feasibility of recovery of mineral resources. Therefore cost of recovery and market price are deterministic factors of the international oil reserve. Therefore, any drastic increase in energy price will tend to increase the cost of the recoverable reserve as well as will require more adoption of complicated technologies. The latter can provide economical recovery options particularly where the developed fields are far out offshore or for those with marginal reservoir. Many oil fields that were considered, not long ago, marginal and/or uneconomical and were not included in the international reserve are increasingly being

763

approaches which resulted in many codes recognizing and placing technical statements and reviewed guidelines. The British standard B.S.7910, the European Committee for Standardisations (CEN) SINTAP and Eurofitnet, APIs RP 579 are all dedicated to address the fitness for purpose (FFS) based assessment of sturctures. These codes and standards provide eminant solutions for the limit state and stress intensity factors for some standard member and sections. Flawed or deformed members requiring a decision based on their level of integrity on whether to repair, replace, rerate or leave requires a dedicated study which has a conservation level inversely proportional to the avialable data and the extent of the analysis. Examples and applications from the media of the available literature survey that cover non-standard or deformed sections have been added to comprehensively test the accuracy of the validity of the theoretical approach. The described cases prove that the utility and the validity of the theoretical approaches are being confidently and remarkably high. 2 STATE OF THE ART Basic knowledge areas for the study that needed to be thoroughly reviewed are: marine risers; fracture mechanics; J-integral method; relevant codes; engineering criticality analysis, Failure Assessment Diagrams (FADs), and; finite element method. A review of these methods is beyond the scope of this work. However, where simple clarifications are a must to communicate an idea are briefly given. 2.1 Acting loads on risers Examples of metalic riser configuration and floaters are illustrated in Figure 1. Risers are normally subjected to the following loads: External and internal pressure; Axial tension/ compression; Vibrations; Temperature; Bending moment (Akram, 2008). These loads are typically resulting from: Applied top tensions. Self weight. Hydrodynamic forces caused by waves and currents. Vessel motion. Riser resonance (Akram, 2008).
(a) The new developed examples of deeper risers,

(b) Some of the recent examples of deeper risers cited to tension rigs
Figure 1. Example of metallic riser configuration and floaters

2.2 Modes of failure of loaded cracked risers Straining rate, fluctuating stresses, stress concentrations, metallurgical flaws, high and low temperatures, corrosion and other special effects can cause an engineering component to fail by fracture which may lead to a catastrophic disaster. The load bearing capacity of a body is reduced by the presence of a crack. As shown in Figure 2(a), the

764

stress is increased locally at the crack tip as a result of a notch effect. There are two ways in which this can lead to permanent failure under Plastic collapse and Fracture. (a) Plastic collapse For every ductile material, increasing the load on the body leads first to yield at the crack tip and the stress then rises locally to the flow stress (usually assumed to be the average of the yield stress and the tensile strength). If the external load is increased further, the plastic zone spreads across the remaining ligament Figure 2(b). Failure by plastic collapse follows when the average stress in the ligament reaches the material flow stress. (b) Fracture Where the material is less ductile, the intense damage is limited to a highly stressed region ahead of the crack tip and failure of the ligament is by propagation of the crack. In a purely elastic material, this is an entirely brittle process with no plastic flow. However, in metal, the high stresses in the crack tip region usually cause yielding to form a local plastic zone Figure 2(c). Nevertheless, when this zone is small, the mode of propagation is still essentially brittle. In less brittle materials, the plastic zone is larger and the facture process becomes one of ductile tearing. This mechanism may persist even where the plastic zone extends across the whole ligament but, in this case, the failure mode may be by plastic collapse as described above.

crack

Figure 3. J-integral contour path

2.3 Review of J-integral method The J-integral provides a means for describing the severity of conditions at a crack tip in a non-linear elastic material, Figure 3. The J contour integral approach is based on the finding that for a two-dimensional crack situation, the sum of the strain energy density and the work terms along a path completely enclosing the crack tip are independent of the path taken. The energy line integral J is defined for either the elastic or elastic-plastic behavior as follows:
J =

w . dy

u T ds x

(1)

Where p = any contour path surrounding the crack tip; w = strain energy density =
w = .d ;
0

T = traction vector defined according to the outward normal n along path p; u = displacement vector; and s = arc length. 2.4 FAD construction using J-integral method A failure assessment diagram (FAD) gives a two parameter approach to assessing a defect. It accounts for the possibility of fracture and plastic collapse separately. These possibilities are plotted on the axes of the FAD as Kr (fracture resistance factor) against Lr (collapse resistance factor). The concept of the FAD requires the use of fracture parameters that cater for large scale plasticity. The j-integral is widely used for this purpose. Once the diagram is generated, an assessment point is considered which may lie in the "acceptable" or "unacceptable" region of the diagram, Figure 4. Several codes of practice provide guidelines for use of the finite element method in creating a failure assessment diagram e.g. R6, API 579, BS7910. The approach is similar in all of these codes:

Figure 2. Modes of Failure of Cracked Structures

765

a. undertake a linear elastic FEA and determine the J-Integral values: Jelas b. carry out a plastic FEA and determine the JIntegral values as a function of load, P: Jtotal=Jelas+Jplas c. determine the reference load of plastic collapse (Pref) d. draw the failure assessment diagram using points (Kr,Lr) where: Kr=(Jelas/Jtotal) Lr = P / Pref e. draw the vertical cut-off on the Lr axis; this depends on the material properties; for some materials: Lr(max) = flow stress / yield stress f. Then assess a point on the FAD and verify the theoretical analysis with the obtained results. 2.5 Different levels of FAD FADs are generally classified in current codes into three different types according to the material properties available for FFS assessment and the conservatism of the diagram; higher-level FADs require more complex data but are less conservative. Level 1 Figure 4(a), a preliminary FAD based on the crack tip opening displacement (CTOD) designcurve method, is the basis of the elasticplastic fracture assessment procedure in BS 7910. Level 2 Figure 4(b) is an alternative FAD based on the lower bound of many curves obtained from experimental data on general austenitic steel. Both level 1 and level 2 contain universal Failure Assessment Line (FAL) which is the criterion line of FAD independent of material properties, as shown in Figure 4. However, level 3 Figure 4(c) is a materialspecific FAD based on the reference stress model. Where the structure did not prove safe after being assessed using a level 3 FAD, a designated material and geometry specific FAD generated using fracture mechanics analysis can be used to assess the structure. A general flow chart of the procedure of a level 3 FAD diagram is shown in Figure 5. 2.6 Flaw Assessment The following is the recommended sequence of operations for carrying out an assessment for a known flaw: a. Identify the flaw type, i.e. planar, non-planer or shape; b. Establish the essential data, relevant to the particular structure;

Figure 4, Schematics of various FADs: (a) level 1, (b) level 2 and (c) level 3 FAD.

c. Determine the size of the flaw; d. Assess possible material damage mechanisms and damage rates; e. Based on the damage rate, assess whether the flaw would grow to this final size within the remaining life of the structure or the in-service inspection interval, by sub-critical crack growth; f. Assess the consequences of failure; g. Carry out sensitivity analysis; h. If the flaw would not grow to the limiting size, including appropriate factors of safety, it is acceptable. Ideally, the safety factors should take account of both the confidence in the assessment and the consequences of failure. i. Determine the limiting size corresponding to the final modes of failure. Appendixe I gives a brief summary of the procedure for unstable crack growth assessment. 3 DISCUSSION AND CASE APPLICATION STUDY Most of the existing recommended practices and standards that address fitness for purpose provide limit state solutions for standard sections and standard joints. The limit state solutions are expressed in terms of empirical formulas that establish FALs of FADs. Needles to say here, that these tend to be conservative solutions. To realize the potential of these methods it is important to seek an alternative solution, which is

766

Figure 6. Dented craked tube model

constructing a finite element model sustaining a 3D. deformation has the extra difficulty of meshing the dented portion of the geometry. 3.4 Validation of dented pipe model Abdel-Nasser et al., had previously extensively investigated modeled dented pipes using finite element method and comparing them to physical experiments results. 3.5 FAD for a dented pipe with circumferential crack
Figure 5. Level 3 FAD flow chart.

advanced assessment using fracture mechanics and finite element methods. The case which was studied here was for a riser joint (pipe) suffering from both a dent and a crack. Consideration for using materials with different tensile strength was also studied. 3.1 Difficulties in FAD construction As summarized in Table 1, different levels of FAD have different requirements with regards to available material data. Lower levels require less data and tend to be more conservative. Nevertheless, the more rationale and optimized options, require more detailed material data as well as material and geometry specific FAD being constructed using a carefully carried out finite element model analysis. 3.2 Validation of plain pipe FAD Burdekin et al., has published work for actual testing carried out on cracked plain pipes. This data was used as reference value to extensively validate the finite element model. The finite element model showed good correlation with the published results. 3.3 difficulty in modeling dented pipe The dent location involves 3D. deformation of pipe geometry. In addition to the normal inherent difficulties involved with 3D. modeling,

Both the validated plain-pipe model and the dented pipe model were combined to produce the model for the dented pipe with a circumferential crack, Figure 6 and 7. Results of the validation analysis are plotted in Figure 8.

Figure 7. Meshed finite elemet model

3.6 Usage of HTS Given two similar design conditions, increasing the grade of steel of line-pipe in simplistic terms will correspondingly decrease the wall thickness and therefore provide cost benefits. In addition to this, a thinner wall thickness will also have various impacts

767

Table 1. Selection of Analysis Levels from Tensile Data. WHEN TO LEVEL DATA NEEDED USE DEFAULT LEVEL Yield or proof strength When no other tensile data available STANDARD LEVELS For quickest Yield or Proof result. Mismatch in 1. Basic Strength: Ultimate properties less than Tensile Strength 10% Allows for mismatch in yield strengths of weld Yield or Proof and base material. Strength: Ultimate 2. Mismatch Use when Tensile Strength. mismatch is Mismatch limit loads greater than 10% of yield or proof strength (optional). More accurate and less 3. SS (StressFull Stress-Strain conservative than strain defined) Curves. levels 1 and 2. Weld mismatch option included. ADVANCED LEVELS Estimates of fracture Allows for loss toughness for crack of constraint in 4. Constraint tip constraint thin sections or Allowance conditions relevant to predominantly those of cracked tensile loadings structure. Needs numerical 5. J-Integral cracked body Analysis analysis As per level 1 but with additional Pressurised information on crack 6. Special components when growth mechanism Case : Leak a conventional and estimation before Break approach does not methods for Analysis indicate sufficient determining crack safety margin. length at breakthrough.

Table 2. Results of F.E. model analysis in tablature format . Standard UM 10 FAD UM 10 Dent UM 10 UM 10 Dent X65 X65 1 1 1 1 1 9.96E1.0008 0.9944 1.00083 9.96E-01 01 1 9.95E0.9758 1.00247 9.95E-01 01 1.0024 9.95E1.0047 0.9297 1.00491 9.95E-01 01 8 9.95E1.0079 0.8106 1.00814 9.95E-01 01 2 0.5723 1.01213 1.00359 1.00348 1.0118 9.60E1.0072 0.3198 01 1.00502 1.00488 7

insufficient data to make a direct like-for-like comparison between, say, X70 and X65 for a given pipe diameter. By increasing the material grade, it is possible to lay pipeline in deeper waters. A thinner wall thickness has a direct impact on this installation method since the requirements for lay barge tensioners are related to the water depth and weight of pipe. 4 RESULTS Figure 8 displays the results of the carried out analysis on tubular joints subjected to both a dent and a crack and comparing them to results for pipes sustaining a crack only. The analysis was repeated for all cases utilizing grade X65 high tensile steel. The same results are presented in tabulated format in Table 2. The analysis is designed around the fracture ratio Kr which in turn is based on stress intensity factor which is a geometry dependent parameter that increases with defect size and the applied stress. Since all the cases were studied to investigate the effect of variation of the material and including a dent in the geometry of the tubular member and identical load ratio Lr stations were used, the produced FADs met the expectations of being closely correlated to each other. However from Table 2, it could be noted that there is a slight decrease in fracture capacity for all the cases containing a dent compared to cases that did not. Also where the X65 material was used, it is noted that there is also a slight decrease in fracture performance probably due to increased yield point and reduced elongation to fracture.

on construction activities. A thinner wall thickness will require less field welding and therefore, in theory, has the potential to reduce construction/ lay time. At present there is
Part Thickness
1.2

0.8

K r

0.6

0.4

0.2

0 0 0.2 0.4 0.6 Lr 0.8 1 1.2 1.4

Figure 8. Results of analysis to validate F.E. model.

768

Figure 9. Results of dented & cracked joint F.E. model analysis.

5 CONCLUSION From this work it can be safely concluded that using standard FAD to assess damage to tubular member containing a dent and a crack is safe and rather conservative. This technique is most valuable when building safety cases based on the proofing of employed non destructive testing method utilized by establishing a FAD for the concerned structure and assesses the smallest detectable crack or the largest crack to escape detection and finding out its potential impact on the structure. Defects revealed during in-service routine inspections that does not endanger the integrity of the structure, need no longer be repaired as with this technique, the behavior of the crack and its stability can easily be modeled. Normally for an in-service dent some yielding must occur, thus changing the material properties in that region. In this study a homogenous material was assumed throughout the model. Future work is suggested to be carried out on accommodating plastic deformation outcome in the material definition in that region. This should not affect the accuracy of this study significantly as behavior of materials possessing higher toughness in general was studied using the X65 model. REFERENCES
Abdel-Nasser, Yehia A. and Rashwan, Ahmad M. 2006. New Simplified Equation for Predicting Ultimate Strength of Damaged Tubular Members, Alexandra University Journal. Annex B of Volume III of the FITNET FFS. API 1993. Recommended Practice for Design, Selection, Operation and Maintenance of Marine Drilling Riser Systems. API RP 16Q. API 1998. Specification for Casing and Tubing. Spec. 5CT: Sixth Edition. Bai, Yong 2001. Pipelines and Risers. Oxford: Elsevier Science Ltd. British Stainless Steel Association website; WWW.bssa.org.uk.

BSI 1999. Guide on Methods for Assessing the Acceptability of Flaws in Metallic Structures. British Standard Institution. BS 7910. Burdeking, F.M. 2001. Experimental validation of the ultimate strength of brace members with circumferential cracks. U.K. HSE Executive Offshore Technology Report, 081. Burdekin, F.M. 2001. The Static Strength of Cracked Joints in Tubular Members. U.K. HSE Executive Offshore Technology Report, 080. DNV 2001. Dynamic Risers. Offshore Standard. DNV-OSF201. El Gammal, M. M. 1975. A new method for estimating fatigue life for ship structures. Shipbuilding, 22: 254. El-Gammal, M. M. 2003. Fatigue Life Prediction in the Presence of Inherited Defects and Corrosion with Marine Applications. Journal of Marine Design and Operations, Proceedings of the Institute of Marine Engineering, Science and Technology. B: 33. Euro FitNet 1999. SINTAP Procedure. Final Version. General Electric Energy website; WWW.gepower.com. Suliman, Akram 2009. Fracture Lifetime Prediction of Marine Riser Systems. PhD. Thesis. The Welding Institute website; WWW.TWI.Org. Walters, D., Thomas, D. and Hatton S. 2000. Design and Optimization of Top Tension Risers for Ultra Deep Water. Proceedings of Conference (1) Floating Production Systems.

APPENDIX I: UNSTABLE CRACK AS FAILURE CRITERIA The condition when the assessed point crosses the FAL can cause the crack growth initiation. Engineering Criticality Analysis (ECA) also considers the unstable crack growth as failure criteria. In evaluating the margin on unstable crack growth, J-Resistance curve is required. The JResistance curve is converted into fracture toughness Kmat vs. crack extension a data. Once the assessed point is outside the safe region, small increment to crack size a, is given. This modifies the assessed point. Because of increase in crack size, K increases while the limit load decreases. The main change is in the value of Kmat, which increases appreciably. Hence the, assessed point has a lower Kr value and a marginally increased Lr value. This process is repeated to check if the assessed point enters the safe region. In such a case, the crack arrest takes place. If the assessed point fails to enter the safe region, the unstable crack growth occurs. The unstable crack growth load is the load at which the locus of the Lr-Kr just touches the FAL. The procedure is breifly summarized in Figure i.1. Consider Figure i.2. The procedure for estimating load to cause unstable crack growth is as follows: 1. The original assessed point is A. 2. The load is increased, the crack growth initiation occurs at point B.

769

3. On further increasing the load, crack growth initiation occurs. Consider point L1. As the stable crack growth takes place, the assessed point is updated based on increased crack size and increased Kmat. The locus followed is L1-L1. The assessed point re-enters the safe region and the crack is arrested. The amount of crack extension done is decided by the availability of J-Resistance data. 4. Now consider the load L3. Here, after exhausting the J-Resistance, still the assessed point is in the unsafe region. The unstable crack growth takes place at this load.

Figure i.2. Unstable crack growth assessment. 5. The objective now is to find a load L2 such

that the locus L2-L2 is tangent to the FAL. This is the limiting load for unstable crack growth.

Figure i.1. Methodology for crack growth analysis.

770

You might also like