Model Summary |
 
|
Model Summary |
|
The riser in this example is a 6’’ production riser, 1080m in length, sited in 705m of water. It is filled with oil and attached to an FPSO. The seabed is modelled as a rigid surface and has a 1.7° slope downwards from PLEM to vessel. The riser analysis is performed here in four stages. An initial static analysis applies gravity and buoyancy forces, and two static restart analyses are then used to introduce vessel offset and current loading. Finally, a dynamic analysis finds the response of the riser to vessel motions and wave loading.
The free hanging catenary configuration represents both the simplest and the cheapest option for the installation and procurement of a flexible riser in deep water. However, the catenary configuration couples the mean, low frequency offset and first order motion of the vessel to the seabed touchdown point response. Where vertical motion of the hangoff point is significant, compression forces can arise at touchdown that, when large, can cause global buckling of the pipe section, significantly reducing the dynamic pipe minimum bend radius. The phenomenon is particularly significant for FPSOs, which tend to have considerable heave response. In addition, the vertical motion is a combination of both the heave and pitch, so catenary risers attached to FPSOs, particular those where the riser hangoff is offset from the vessel centre of gravity, are likely to experience compressive loading and possible buckling.
The issue of modelling compression in flexibles was the subject of the publication McCann et al., (2003). The example data used here is based on one of the examples described by McCann et al, and the results presented here are a small subset of the results presented in the paper. Interested readers are referred to the relevant proceedings for a full version of this material.
McCann et al list a number of conclusions, some of which are repeated because of their significance to this particular example.
1.Flexcom can accurately predict both the onset of buckling and post-buckling behaviour, provided sensible values of element length and time-step are used, and provided the Euler buckling within each individual element is not exceeded.
McCann et al demonstrate this by comparing Flexcom output for a column instability analysis with analytical and third-party solutions. Complete agreement with independent solutions is reported.
2.The paper proposes a non-dimensional velocity parameter (VHangoff/VTerminal) which is both a simple and useful measure for assessing the likelihood of compressive buckling. VHangoff is the maximum heave velocity of the riser hang-off, whereas VTerminal is a riser “terminal velocity” as defined by McCann et al. Where this parameter is greater than 1, buckling is expected. Further details are contained in McCann et al., (2003). For the riser and vessel data used in this example, VHangoff is 4m/s and VTerminal is 2.72m/s. So the ratio is 1.47, making compression likely according to the McCann et al criterion. That compression does indeed occur is demonstrated in the dynamic analysis results.
3.Accurate modelling of compression is sensitive to element length and dynamic analysis time-step. McCann et al demonstrated this with a range of sensitivity analyses. In this example, the effect of time-step size is investigated.
4.Large variation between results from deterministic analysis can exist when buckling occurs. McCann et al recommend that a stochastic approach be used to extrapolate extreme curvature based on the standard deviation of curvature in the touchdown region, even in the case of regular wave analysis. In the dynamic analyses reported here, both max/min and standard deviation plots are presented to illustrate this recommendation.