Feasibility of curved pipe pull operations

Abstract

Due to often existing subsea infrastructure or challenging seabed conditions, engineers are forced to design newnearshore pipeline trajectories that are not always in a straight line. Since subsea tie-in operations are often complex and relatively expensive, alternative methods like pulling pipelines into a curve are worthwhile. By means of this research fundamental knowledge of pipe pull operations into curved trajectories is acquired. This thesis research focuses on the soil - structure interaction between the sliding pipeline and the seabed. A more thorough understanding is obtained of the geotechnical processes that create lateral resistance of the pipeline against the forces applied by the installation vessel. The main goal behind this thesis is to expand the capabilities of pipeline installation by exploring the feasibility of curved pipe pull projects. The research is based on three fundamental elements: 1. An extensive theoretical study 2. A validated numerical soil-structure interaction model 3. A successful physical test campaign of thirty-two experiments The theoretical study of the thesis focuses on the physical processes that play a role during curved pipe pull operations. Research is conducted to gather knowledge of the three main subjects of research, being: 1. Structural behaviour of a concrete covered pipeline 2. Geotechnical behaviour of soil under lateral pipeline displacement 3. Geotechnical behaviour of soil under axial pipeline displacement Since this was the first known research into curved pipe pull operations, influences of dynamic environmental loads are excluded from the scope of research. The second stage of the research focused on creating an engineering model that enabled to predict to what extend partially embedded pipelines can be pulled into a curved trajectory on sandy seabeds. After considering multiple simpler structural models, the final result was a non-linear spring supported tensioned bending beam model. Within this computational prediction model: the lateral soil resistance is modeled by means of a P(Y) spring model, while the pipeline is represented by the tensioned bending beam. The applied lateral soil resistance model is computed based on researches ofWang et al. (2018) and Verley and Sotberg (1994). After themodel was completed, a purpose made test facility was created to pull scaled pipelines into a curved trajectory. During the test-campaign, five different model pipelines were tested on a scale range of 1/11.8 to 1/5.3. The physical test campaign provided valuable experimental data of thirty-two successful drained curved pipe pull tests. By means of the acquired set of experimental data the model was able to be validated. Since the model pipelines differed in diameter, specific weight and bending stiffness, the influence of those parameters is examined. From the fact that the majority of the model pipelines was pulled into a radius that was 0-30% lower than the computed radius,we can state that the numerical model predictions gives awell defined upper limit of the controlled pull radius of pipelines. Along the classified critical curved pulls, the imposed lateral force during the experimentswas within 20% of the maximum predicted force in 86% of the physical tests. From this observation we can conclude that the numerical model captures the maximum lateral pull force with a high accuracy.