Introduction |
 
|
Introduction |
|
This example considers the analysis of a dual barrier production riser, which is modelled as three concentric tubes using the Flexcom pipe-in-pipe modelling facility. The model comprises a 12¾” OD outer riser, an 85/8” OD inner riser and 4½” OD production tubing; each has a thickness of ½”. The outer and inner risers extend from the spar deck to the mudline, while the tubing extends from the deck to a packer 8000 ft below the mudline (the water depth is 4000 ft). The annulus between outer and inner risers contains fluid with a density of 9 ppg, while the inner riser/tubing annulus has fluid with a density of 2 ppg. The fluid in the tubing itself is 7 ppg.
The example is sub-divided into two main parts:
•Standard analysis: A static analysis of the riser subjected to current loads and vessel offset, followed by a dynamic analysis including wave loading and first-order vessel motions.
•Specialised case: A series of analysis stages culminating in the fatigue assessment of the inner riser subject to current-induced vortex induced vibration (VIV).
The main purpose of the example is to demonstrate the program options for analysing pipe-in-pipe systems. There are two main aspects to pipe-in-pipe modelling in Flexcom, both of which are invoked here. The first involves identifying which sections of the model are contained within which other sections (i.e. Pipe-in-Pipe Sections). So in the present example you specify that the tubing above the mudline is contained within the inner riser, and that the inner riser is everywhere contained within the outer riser. The second modelling aspect involves specifying pipe-in-pipe connections between nodes on the respective sections. Linear and non-linear connections are available, and these can be used in combination, as they are here. Linear connections are used at regular intervals to model spacers or centralisers. Elsewhere non-linear connections are specified: these have a very low stiffness when the sections are in their respective undeformed positions, but the stiffness increases exponentially as the gap between sections approaches zero, to provide a simple contact model.
The example also showcases a modelling approach for estimating fatigue damage on inner pipes induced by VIV. This is a specialised methodology in which the VIV motions of the outer pipe, as estimated by a modal analysis and Shear7, are applied to the inner pipe using fabricated sinusoidal motion time histories, and then dynamic fatigue is computed in the usual manner via rainflow cycle counting. Refer to VIV Induced Fatigue of Pipe-in-Pipe Systems for further details.
This example system is intended to represent a production riser operating on a spar. However since the main purpose is to demonstrate pipe-in-pipe modelling, a number of aspects of spar riser analysis (particularly those illustrated in Example A02 - Spar Production Riser) are only very simply modelled here. In a realistic spar production riser model most or all of these simplifications would be addressed.
•Although the tubing is modelled to 8000 ft below the mudline, soil/tubing interaction is not explicitly modelled using P-y curves. Instead a simple lateral (DOF 2) restraint is applied at every 100 ft between packer and mudline.
•Likewise interaction (contact) between the outer riser and guides in the spar hull is not explicitly modelled - again a simple lateral restraint is applied at 100ft intervals within the hull.
•A keel joint is not included in the model.
•The top tension is modelled as a simple vertical point load.
•As current-induced VIV is the focus, the tubing below the mudline is omitted.
•Contact between the outer riser and guides in the spar hull is simulated using linear node springs, rather than attempting to use contact surfaces. As the modal analysis is inherently linear by nature, any non-linear effects caused by intermittent contact cannot be captured in the VIV assessment. In order to ensure conservative results, node springs of relatively high stiffness are used, as this will cause elevated curvatures to appear in the region of the lowest riser guide on the spar hull (i.e. at the intersection between restrained and unrestrained sections). Defining the rigidity/flexibility of these linearised connections is a matter of engineering judgment. If you re-run the simulation using different input values, you will be able to examine the sensitivity of fatigue damage estimates to this parameter.
•Only one current profile is considered in Shear7, whereas in reality a range of current profiles would be simulated to cover the range of loading conditions likely to be experienced by the riser system.