Elsevier

Applied Ocean Research

Volume 48, October 2014, Pages 292-300
Applied Ocean Research

Refined analytical models for pipe-lay on elasto-plastic seabed

https://doi.org/10.1016/j.apor.2014.10.003Get rights and content

Highlights

  • We provided a refined analytical model to estimate the pipe–soil interaction on elasto-plastic seabed during pipe-lay.

  • We explored the effect of the rebound modulus.

  • We improved the accuracy of the solution by using H as the constant axial force.

Abstract

This paper presents a refined analytical solution for pipe-laying on an elasto-plastic seabed. The solution builds on previous work, extending it to include elastic rebound of the pipe. The approach simplifies the pipeline as the combination of four segments: a natural catenary through most of the water column; a boundary-layer segment in the water close to the seabed where the bending stiffness of the pipe modifies the shape of the catenary; a beam under uniform tension through the touch down zone up to the point where maximum pipe–soil load concentration occurs, within which the soil responds plastically; and finally a rebound segment, also modelled as a beam under uniform tension, where the soil rebounds elastically as the pipe–soil contact force reduces back to the submerged pipe weight. Continuity of displacement, gradient, bending moment, shear and deduced tension along the pipeline are preserved. In comparison with previous models, such as a rigid-plastic seabed model, the distribution of seabed resistance is continuous. Results from the solution are presented for the case of seabed resistance to pipe penetration increasingly proportionally with depth, as is approximately the case at very shallow depths within the seabed. The lower the gradient of seabed resistance, the greater is the pipe embedment, but the maximum contact force and curvature of the pipe both reduce. Analyses also show that the rebound stiffness can have a marked effect on pipeline embedment, which increases with increasing rebound stiffness. However, the effect on pipe embedment becomes small beyond a certain ratio of rebound stiffness to shear strength.

Introduction

Pipe–soil interaction is of great importance during pipeline installation, as it influences the pipeline shape and internal force distribution, as well as the pipeline embedment and stability during subsequent operations. Analytical solutions are complicated by the need to consider non-linear, plastic response of the seabed, and to allow for the effects of cyclic remoulding of the soil due to the dynamic pipe motions in the touchdown zone [1], [2]. The earliest solutions for pipe-lay assumed the seabed as rigid, which greatly simplified the pipe–soil interaction problem [3], [4]. Such solutions exaggerate the maximum pipe–soil contact force and the curvature at the point of contact. Later, solutions based on linear elastic seabed response were developed [5], [6], [7]. These solutions, in particular the closed form solution and analysis framework of Lenci and Callegari [5], were relatively simple and easy to implement, and represented a significant advance. From these solutions it is possible to obtain reasonably accurate estimates of the pipe shape and distribution of pipe–soil contact force through the touchdown zone (TDZ). Preliminary estimates of pipe embedment could be made by adopting values of secant stiffness that reflected the non-linear response of seabed sediments, although the solutions could not capture the pipeline profile in the seabed beyond the TDZ.

As the exploitation of oil and gas has moved from shallow to deep water, where the seabed typically comprises soft, fine-grained sediments, the importance of capturing the plastic response of the seabed has increased. Palmer [8] suggested an analytical pipe–soil interaction model for pipelines on rigid-plastic seabeds. Although certain simplifications were adopted in the model, including uniformly distributed soil resistance and limited consideration of the overall pipeline shape, it represented an important step towards improved analytical modelling of pipe–soil interaction in the TDZ. Wang et al. [9] and Yuan et al. [10] combined the models of Lenci and Callegari [5] and Palmer [8] to establish an analytical solution for the complete pipeline for a rigid-plastic seabed where the pipe-soil resistance was assumed to increase proportionally with pipeline embedment. This model explored the effects of cyclic remoulding of soil in the TDZ, as suggested by Westgate et al. [1], by considering softened soil resistance profiles. The main limitation was that, in order to simplify the solution process, the rebound compliance of the seabed was neglected. This resulted in a discontinuity in the soil resistance, as detailed later. The solution presented here has improved the analytical model of Yuan et al. [10] by considering the rebound stiffness of the seabed, thus preserving continuity of soil resistance. It has also considered a different assumption for the axial force in the pipe segments close to the seabed, as discussed in the following section.

Section snippets

Analytical solutions

The analytical boundary layer solution of Lenci and Callegari [5], on which the present work is based, adopted an assumption of constant axial tension in the segments close to the seabed, with the tension fixed at the value (T) at the transition to the catenary section (point P1 in Fig. 1). The assumption is reasonable, since the angle of the pipeline to the seabed is relatively small there, but it leads eventually to a tension in the near-horizontal pipeline along the seabed that is

Solution

No explicit solution can be obtained for Eq. (36). In the present study, we have adopted a numerical method, similar to that of Yuan et al. [10]. For a constant axial force of H, the solution is shown schematically in Fig. 5. The value of H estimated from Eq. (35) is taken as the initial value. Meanwhile, a two dimensional (X2, X3) space is established with X2lower < X2 < X2upper, X3lower < X3 < X3upper, where the subscripts “lower” and “upper” represent the lower and upper bound respectively. Then, the

Soil properties

In moderate to deep water, the shear strength profile may be approximate as increasing proportionally with depth with a mudline intercept of zero. The strength gradient will typically lie in the range 1–2 kPa/m for normally or lightly overconsolidated fine-grained sediments. Where a surface crust exists, which is common [15], [16], much higher strength gradients pertain in the zone of interest for pipeline penetration, as illustrated in Fig. 6. Cyclic pipeline motions during laying will remould

Validation and comparison

In order to illustrate the above model with some examples, a typical pipeline is selected with elastic modulus E = 210 GPa, outer and inner diameters of the pipeline d1 = 0.6 m and d2 = 0.55 m. The density of steel pipe is ρs = 7.85 × 103 kg/m3, inclination slope at P1 is φ0 = 80°, the density and depth of sea water are ρw = 1.03 × 103 kg/m3 and Y0 = 1000 m, and the intact shear strength gradient is ksu = 12 kPa/m. Two sets of comparison examples with St = 1 and 8 are made to validate the present models and quantify the

Influence of soil sensitivity

The influence of the seabed stiffness is explored for an initial strength gradient of ksu = 12 kPa/m, and sensitivities ranging from 1 to 8, so that the operative strength gradient ranges from ksu = 12 kPa/m down to ksu = 1.5 kPa/m. The rebound stiffness is maintained at kr = 100kp = 375ksu. The results are shown in Fig. 10. The maximum embedment increases by a factor of 5.5, from 0.12 m to 0.66 m, as the operative strength gradient is reduced from 12 to 1.5 kPa/m. Correspondingly, the maximum contact force R

Influence of rebound stiffness

The effect of the rebound stiffness is explored in Fig. 11, for six different rebound stiffnesses of: kr = 10kp, kr = 20kp, kr = 40kp, kr = 60kp, kr = 80kp and kr = 100kp. The operative soil strength gradient was taken as ksu = 1.5 kPa/m, so kp = 5.64 kPa. The maximum embedment of the pipe exceeds the final embedment in accordance with the rebound stiffness. Also, the greater the rebound stiffness ratio, kr/kp, the greater is both the maximum and the final embedment. For rebound stiffness ratios of 60 or more,

Conclusions

This paper has presented an analytical model for pipe-lay on an elasto-plastic seabed. The analysis improved previous analytical work of Yuan et al. [10] by considering rebound of the pipe beyond the point of maximum contact force (and embedment). In addition, an alternative assumption of constant tension was adopted in solving the differential equations within the boundary layer, touchdown and rebound segments. The constant tension was taken as H, the horizontal component of tension in the

Acknowledgement

The authors would like to acknowledge the support of the Grant No. 2014T70574 from Postdoctoral Science Foundation of China, China, and the Grant No. 51409228 from the National Natural Science Foundation of China.

References (22)

  • L.Z. Wang et al.

    Numerical analysis of pipeline in J-lay problem

    J Zhejiang Univ Sci A

    (2010)
  • Cited by (15)

    • A numerical model for geometrically nonlinear analysis of a pipe-lay on a rough seafloor

      2022, Ocean Engineering
      Citation Excerpt :

      Therefore, to overcome the use of GUI of the abovementioned commercial software, at least through the course of the design process where the overall pipe-lay configuration is of interest, in recent years, plenty of various in-plane techniques for simplified pipe-lay in-plane calculations have been proposed. However, unfortunately, these techniques, both analytical (e.g. (Lenci and Callegari, 2005; Yuan et al., 2014, 2012)) and numerical (e.g., (Wang et al., 2010a; Zan et al., 2016)), based on equilibrium considerations, usually consider pipelines by separating them into segments according to the discontinuities in the loading conditions or cross-sectional geometry, enforcing idealized structural behavior for each segment and adopting simplifying assumptions regarding the continuities between the segments. Those constraints not only lead to the reduction in accuracy but also make the above techniques very case-specific and consequently unfeasible for the calculation of internal force envelope diagrams, where continuously changing seafloor geometry must be considered.

    • Static analysis of offshore pipe-lay on flat inelastic seabed

      2020, Ocean Engineering
      Citation Excerpt :

      Consequently, the above penetration-resistance relations have been also adopted in some simpler 2D models, whose aim is to provide a less time-consuming alternative, which avoids the use of cumbersome GUI inherent to the above 3D software, considering that the dominant first order effects, which define a general pipe-lay configuration, are usually static and occur in-plane (Lenci and Callegari, 2005). In the recent decades there was a variety of such models suggested, both analytical (e.g. (Lenci and Callegari, 2005; Yuan et al., 2014, 2012)) and numerical (e.g. (Wang et al., 2015; L.-Z. Wang et al., 2010; Zan et al., 2016)). Those models are usually based on a division of the pipeline into a number of segments, within each one idealizing its particular structural behavior.

    View all citing articles on Scopus
    View full text