Significance of Active Tension Compensation during S-lay Pipeline Installation

Abstract

During S-lay pipeline installation, tension fluctuations in the suspended pipeline are compensated by the tensioner system on the vessel. The goal of this thesis is to answer the questions: to what extent does the active tension compensation (ATC) influence the pipeline integrity during offshore installation. Secondly, which tensioner model should be used for pipelay analyses that can give accurate results? And lastly, what is the implication on the existing tensioner models (deadband and linear damping with velocity cap) that are currently used in pipelay analyses?

The pipelay operation is modelled in OrcaFlex with 7 case studies from shallow to deep water based on full-scale measurement data measured during offshore installation. For each case study, three types of measurement data are available: vessel motions, tensioner pay-out motion and pipeline top tension. Vessel motions and tensioner motion signals are used as inputs to the OrcaFlex model, whereas the measured tension signal is used for comparison with the output pipeline top tension for model validation.

It is observed that the ATC becomes more important in deeper water as the reduction in maximum von Mises strain in the sagbend and accumulated fatigue damage become more significant with increasing water depth. The maximum von Mises strain is found to be in between 0% and 2.2% lower from the model with uncompensated tension. Moreover, the reduction in accumulated fatigue damage is more significant as the allowable standby time increases between 3.2% and 19.5%.

Subsequently, different tensioner models are included in the simulation. The tensioner model computes the pay-out motion based on the pipeline top tension deviation from a set tension at each time step. The initial model with specified tensioner motion is then used as a benchmark for the other tensioner models. It is found that the existing tensioner models are not conservative as the maximum von Mises strain and accumulated fatigue damage are underestimated, with the exception of the accumulated fatigue damage in shallow water cases.

As an alternative to the existing tensioner model, a tensioner model with a PI controller is introduced to represent the PI controller of the tensioners systems on the vessel. This PI controller model scripted in Python is implemented as an external function to the OrcaFlex model.

The controller gains are tuned proportional to the logged controller gains from the vessel so that the tensioner model produces the pay-out signal similar to the measured pay-out. Based on the pipeline integrity check, the tensioner model with PI controller is valid and can be used for a more accurate investigation of the pipeline integrity in all cases.