The development of offshore fields has reached large water depths, and hence more challenges are met in design of offshore systems. The large water depths challenge the designer in terms of high temperature and pressure in complex reservoirs in addition to extreme environmental conditions. Pipelines and risers are normally designed for a service life of 20–25 years. For some oil fields, the average service life of a riser is only 50% of its planned service life. Hence, it is important to perform extensive global and local analyses to provide a desirable service life of the riser system. One of the main topics that should be evaluated thoroughly is the fatigue damage of the riser. By performing local fatigue analyses, the fatigue life can be determined and so also a desirable service life.
Fatigue is mainly a condition of structural damage that occurs as a result of a cyclic loading. When looking at fatigue of flexible risers, the cyclic loading is initiated by environmental loads causing movement of the floater and hence induces dynamic loading on the riser. The intersection between the riser and the floater is then one of the most vulnerable area for fatigue, depending on floater type, riser system, etc. When evaluating fatigue life of a flexible unbonded riser, the expected life of the pipe structural layers is found. The main focus is on the tensile armor and pressure armor layers, as they are generally exposed to fatigue failure and known to limit the service life of the pipe. As of today, there are no recognized, well-defined mathematical methods to calculate aging of the plastic layers in the cross-section. Hence; this is not discussed further in the thesis.
Typically, the hang-off area must be protected from over-bending so additional devices are added. A common choice is to use a bend stiffener, which is a device that contributes to a moment transition between the riser and the end connection on the floater. The most common approach for evaluation of the fatigue life of a riser, is by making use of S–N curves and Minor summation method. Each layer of the riser cross-section has its own material properties, and hence its own S–N curve. The Minor summation method provides a failure criterion for the fatigue damage, and this should be fulfilled to gain a suitable riser configuration. A basic procedure of fatigue assessment is provided by DNV, DNV-RP-F204:
- Define fatigue loading
- Identify locations to be assessed
- Global riser fatigue analysis
- Local stress analysis
- Identify fatigue strength data
- Fatigue analysis
- Further actions if too short fatigue life
Global analyses are performed to establish extreme interface loads during operation and long term distribution of fatigue loads for the deep-water riser application. In the global extreme analysis carried out in this thesis, it is found that near position of the vessel contribute to the highest maximum tension in addition to compression of the riser at touch down point, TDP. The far position normally cause highest value of top tension since this is where the riser is stretched the most. The curvature obtained in the analysis show a high result at the TDP, which has been found unacceptable in comparison with the minimum-bending radius, MBR. Even though the curvature result is unacceptable, it may be a result of numerical errors in the computations. Thus, the extreme analysis should be evaluated more thoroughly and optimizations of the riser configuration should be carried out.
The highest and lowest results of tension in combination with bending angle found from the global extreme analysis are applied for the bend stiffener design. In the bend stiffener design, maximum shear force is applied. The maximum shear force is found for far position of the vessel. When the initial design of the bend stiffener is found by applying the shear force with accompanying bending angle, the design should then be verified by local analysis. Two cases are evaluated, where maximum bending angle in combination with maximum and minimum tension for the specific load case are applied. The verification has been done in terms of curvature and stress distribution along the riser in the bend stiffener area. The curvature is distributed with small values at the root of the riser and at the intersection between the bend stiffener and the riser. All stresses found for the riser cross-section are acceptable with reference to design criterion given by ISO 13628-2.
The global analysis performed to establish long term distribution of fatigue loads, provides time series of maximum top tension and bending angles for the riser. The results obtained are applied in fatigue analysis of the structural layers of the riser, with focus on the bend stiffener area and the first tensile armor layer. The fatigue analysis is based on longitudinal failure mode, with evaluation in reference to the failure criterion of fatigue damage found in DNV-RP-F204. Sixteen different cases, with different combinations of wave height, period and direction, are assessed. Maximum fatigue damage was found for the same node in each case, and evaluated based on a safety-factor of 10. By summation of the individual fatigue damage the total fatigue damage was found to exceed the damage criterion. Even though the criterion is not fulfilled, it is important to know that no optimization of the configuration has been carried out.
A parametric study has been performed to outline the effect on stress level in the bend stiffener area. Temperature variation in the annulus of the riser and change of material properties has been assessed. The variation of temperature in the annulus of the riser resulted in a small increase of the stresses in the structural layers of the cross-section. However, the increased stresses obtained did not exceed the design criterion for the tensile armor wire. Fatigue damage has been evaluated for change of temperature. An overall increase of the fatigue damage was found, and as for the base case, the failure criterion was exceeded. The change from non-linear to linear material resulted in a decrease of the stress level in the bend stiffener area. Fatigue damage was not evaluated for this case.