Fatigue And Reliability Study Of Deepwater Steel Catenary Riser Near Touchdown Point

Riser is a slender structure connected from seabed to platform, and mainly usedfor drilling and oil&gas transportation. Among all riser types, steel catenary riser(SCR) is of easy installation, low cost and compatibility with more platform types.Therefore, its use has extensively increased in recent years. Although SCR istechnically feasible, the fatigue damage near touchdown point (TDP) due to itsparticular configuration is still one of the most challenging issues in offshore industry.If the environment loads can be fully taken into account, and the SCR-soil interactioncan be modeled reasonably, the prediction accuracy of the fatigue damage near TDPwill be improved. This can greatly decrease the cost of the riser manufacture.As an important part of the projects of “The probability prediction approach studyof the VIV induced deepwater riser fatigue damage based on the ocean currentrandom characteristics” granted by National Natural Science Foundation of China,“Deepwater subsea pipe and riser technology” granted by Major Scientific andTechnological Project of China and “Multi-modes VIV induced fatigue study ofdeepwater riser” granted by National Key Laboratory Self-Research Project, thisdissertation is aimed to study the fatigue damage characteristics of SCRs near TDP,and develop reasonable and effective assessment system.Based on the International latest progress of the fatigue damage study for SCRs,this dissertation intensively studied the SCR-soil interaction, frequency domain andtime domain prediction technology for VIV, fatigue performance of SCRs near TDPand reliability prediction, and some positive results are obtained. The main works ofthis dissertation are as follows:(1) An overview of research progress about the platform system coupled responseprediction, SCR-soil interaction simulation, VIV prediction, riser fatigue damageand reliability analysis is made, and some aspects with research gap or necessityto be improved are presented. Meanwhile, some important pipe-soil interactiontests and VIV test are in detail summarized.(2) According to the frequency-domain/time-domain transformation theory, a hullelement with single node is developed by using frequency domain hydrodynamiccoefficients. The three-nodes truss element and beam element considering thegeometry nonlinearity are employed to simulate the mooring line and riserrespectively. The hull element and truss/beam element are connected by a newdeveloped multi-point constrain. Next, a full coupled prediction model forhull/mooring/riser system is constructed to insure that the coupled response iscalculated in an assembled matrix.(3) A code for the SCR-soil interaction during initial penetration is developed basedon beam-spring model. It is used to study the effect of the maximum penetration depth, seabed stiffness and riser weight on the riser shape near touchdown zone(TDZ). The existing seabed trench models and their faults are discussed, and thena new model is proposed based on the observed trench shape in Gulf of Mexico.The phenomena obtained from SCR-soil interaction test are introduced, and alinear hysteretic model for the vertical SCR-soil interaction is proposed bysegmentally linearizing the interaction process. The lateral interaction issimulated by spring model with the stiffness as a function of the penetration depth.According to the trench model, vertical and lateral interaction model, atouchdown element associated with SCR is developed, and then the qualitativeanalysis is carried out to demonstrate that the element can well capture themobilization and release of the clay suction.(4) The consistency of the lock-in region obtained from the forced vibration test andfree excited vibration test is analyzed. A principle for dealing with the power-insuperposition region between two excited modes VIV is proposed, and then across flow (CF) and in line (IL) VIV prediction model in frequency domain isdeveloped based on the force vibration test data. Several VIV test models aresimulated by this model, and the obtained displacement and excited modes arecompared with the test data. The effect of the existence of TDZ on the response atTDP is studied, and the unreasonable results obtained from traditional truncatedVIV prediction model are demonstrated.(5) Considering the intrinsic defect of the frequency domain VIV prediction model,this study develops a time domain prediction model based on a proposed lock-incriterion. This model is validated by simulating the top tensioned riser (TTR) andSCR VIV test models. The mode shape direction of SCRs is discussed, and themethod that the frequency domain and time domain VIV prediction model use the2D modes to replace3D modes is validated. By using the touchdown element inthe time domain VIV prediction model, the effect of the seabed parameters on theexcited modes near TDP and the global displacement is studied.(6) A fatigue damage calculation code is developed based on rain flow countingmethodology. By comparing the numerical and test data of a VIV test model, thecode is validated. As regards to the frequency domain model, three approachesfor the multi-modes VIV induced fatigue damage are discussed, and differentapproaches are recommended for the CF and IL VIV respectively. The wave andVIV induced responses of a SCR near TDP are carried out in time domain, andthe sensitivity of fatigue damage near TDP to the wave and seabed parameters isdiscussed. The frequency domain model and time domain model obtained fatiguedamages are compared to further prove the recommended approaches for thefatigue damage calculation in frequency domain.(7) The response surface method, Monte-Carlo method and Rosenblueth method areintroduced. Combined with the frequency domain VIV prediction approach and platform system coupled model, a code is developed for the calculation of theriser fatigue reliability and fatigue safety factor (FSF). Next, the probabilitydistribution type of VIV induced fatigue damage, the sensitivity of the fatiguefailure probability to uncertainties and the relationship between the coupledresponse and VIV induced annual fatigue failure probability and FSF are fullstudied.