Progress Reports :June 2006 December 2005 June 2005 December 2004
SEAFLOOR INTERACTION WITH STEEL CATENARY RISERS
OBJECTIVE: The research will improve the current state of understanding regarding the basic mechanisms affecting the seafloor stiffness within the Steel Catenary Riser (SCR) touchdown zone, and (2) develop a means for providing quantitative estimates of seafloor stiffness and damping and their variation over the life of the project for various soil and site conditions.
APPROACH: The geotechnical studies in this project will use finite element analyses to develop load deformation relationships to describe soil-pipe interactions for various conditions of soil properties, trench conditions, and characteristics of riser motions. The hydrodynamic studies will investigate riser-structure-fluid interactions for various sea states, floating system characteristics, and seafloor conditions. The parallel studies will be coordinated throughout the project to ensure proper accounting of interaction effects.
BENEFITS TO MMS & INDUSTRY: Results will allow a more accurate estimation of SCR fatigue life in the touchdown area.
DEPLOYMENT OF RESULTS: The studies described above will be documented in a final report. Guidelines and recommendations will be provided for appropriate seafloor boundary conditions (stiffness, damping) for use in fatigue stress analyses of steel catenary risers for a practical range of seafloor conditions likely to be encountered in the Gulf of Mexico.
PROJECT ORGANIZATION & TIMING: Phase 1 of this project has provided an overall study framework and plan, validation of FEM tools for use in this project, and an initial identification and assessment of the important parameters for modeling SCR’s and developing boundary conditions in the touchdown area for use in estimating the fatigue life of SCR’s. This Phase 2 will include model development, FEM calibration, and parametric studies for various seafloor and loading conditions. Phase 3 will evaluate model predictions in light of laboratory measurements and field observations, recalibration/refinement of the model as necessary, and formulation of final guidelines.
ANTICIPATED NUMBER OF PHASES: 3
PROJECT PLAN FOR PHASE 2 (2005-2006):
Scope and Plan: Develop a numerical model, perform parametric studies, and identify conditions most influential on the magnitude of bending stresses in the riser pipe. Parameters to be evaluated included soil strength and stiffness, trench geometry, the effects of trench collapse and infilling, strain rate effects, and non-linear effects associated with a range of amplitudes and velocities of riser motions. The range of riser motions near the touchdown zone will be approximately estimated using a numerical code accounting for the interaction among a floating structure, its riser/mooring system and the seafloor.
Soil and Riser Interaction
Model Framework - In principle, the soil-riser interaction is a three-dimensional problem involving a beam element embedded in a continuum (Figure 1a). Analysis of the full three-dimensional problem is within current computational capabilities, albeit with a high level of computational effort. Given that in the early stages of the proposed research the focus is to be on understanding basic interaction mechanisms rather than high levels of numerical accuracy, the initial series of studies will employ a simplified model comprised of a pipe support on a bed of springs as shown in Figure 1b. Two-dimensional plane strain finite element analyses of a circular pipe embedded in a continuum (Figure 2a) will be used to develop load-deformation (P-d) relationships illustrated conceptually in Figure 2b. The computational efficiency gained from this approach will permit evaluation of P-d relations for a wide range of trench geometry and soil conditions. At a later stage in this study, true three-dimensional analyses will be performed to evaluate the accuracy of the simplified analysis described above.
The spring supports for the pipe will not in general be uniform. For example, at the initial point of contact between the riser and the seafloor the trench will be very shallow. Further, the high amplitude of cyclic pipe motions at this point will likely cause severe remolding and softening of the soil surrounding the pipe. In contrast, near the mid-point of the touchdown zone (Figure 1a) the trench depth will be a maximum and the soil will be essentially intact; hence, the seabed will be relatively stiff in this region. Bending stresses occurring in an elastic pipe resting on non-linear, non-uniform spring supports will be computed in a separate series of finite element analyses.
Parametric Studies of P-d Curves - Analytical studies will be conducted to characterize the effects of various soil and site variables on the P-d curves describing the relationship between pipe deflection and soil resistance. These studies will be based on the two-dimensional finite element model of a pipe embedded in a continuum discussed earlier (Figure 2a). The following variables will be considered:
Soil strength Soil strength will clearly be a primary parameter influencing the resistance of the seafloor to riser pipe motions. A complicating factor is that repeated load cycles will tend to degrade the strength of seafloor soils below their intact values. The issue of degradation raises two questions: (1) to what level will the soil strength degrade, (2) and what will be the likely areal extent of strength degradation. Knowledge of soil sensitivity, the ratio of intact to remolded soil strength, can provide a reasonable estimate of the degree of strength degradation. Existing databases of seafloor soil properties, including sensitivity, can provide a start toward establishing reasonable ranges of strength degradation. Since strength loss is related to cumulative straining cycles experienced by the soil, finite element studies of the strain distribution surrounding a loaded pipe can provide reasonable first estimates of the areal extent of strength loss.
Trench geometry Riser embedment conditions in the touchdown zone vary from zero at the touchdown point to 4-5 pipe diameters in the middle of the touchdown zone. The effective soil stiffness beneath a pipe in a trench will be considerably greater than that of a pipe resting on the seafloor due to the increased level of confinement. Offsetting this effect somewhat will be the tendency of the trench to widen as a result of lateral riser motions. At sufficiently great trench widths, the riser pipe effectively rests on a free surface, and the effect of trench depth becomes negligible. Finite element analyses for various combinations of trench depth and width will be performed to characterize the influence of trench geometry on P-d curves.
Trench infilling ROV examinations of SCR’s in the touchdown zone often show areas where the trench collapses, effectively filling the trench with remolded soil. The effective seafloor stiffness for the condition of an infilled trench will be considerably greater than that for an open trench; hence, trench collapse is a scenario that definitely should be considered. Simulation of the trench formation and collapse process is exceedingly complicated. However, a first-order assessment of the effect will be attempted through finite element simulations of a pipe embedded in a trench infilled with remolded soil.
Rate effects The dependence of clay undrained stress-strain-strength behavior on rate of straining is well established, with the stiffness and strength of clays increasing with increasing strain rate. Strain rates in the soil surrounding a riser pipe are non-uniform; i.e., they are greatest near the pipe boundary and decline with increasing distance from the pipe. Evaluation of a field of non-uniform strain rates on the net soil resistance beneath the riser pipe is possible from finite element analyses of pipe loading in a rate-dependent soil. The velocity of the riser pipe controls the overall magnitude of strain rates within the soil mass; therefore, the P-d curve for a soil-riser pipe system must be considered to be a function of velocity of the riser pipe.
Pipe pullout In the immediate vicinity of the touchdown point, the cyclic riser motions will generally be such that the riser pulls away from the seafloor during the upward movement cycles. Similarly, during the downward cycles, there is no soil resistance to pipe motions until the pipe comes into contact with the seafloor. The touchdown point is typically a critical zone with regard to pipe bending stresses. Therefore, close attention to the details of soil resistance is likely to be crucial in this region. Accordingly, P-d curves simulating the change in the contact condition will be developed for soil conditions at the contact point.
Riser Analysis - The amplitudes, velocities, and frequencies of SCR motions near and at the seafloor influence the depend upon the motion characteristics of the Floating Production System (FPS) from which the SCR is suspended, and the wind, wave, and current environments affecting the FPS and the SCR. For example, the heave for TLP’s and Spars are generally much smaller than a semisubmersible, and spars generally have larger pitch and roll than a TLP or semisubmersible. During a hurricane, SCR motions are influenced by waves and near surface currents, whereas during loop currents SCR motions are influenced by currents penetrating to considerable depths and resulting vortex-induced vibrations (VIV).
In this study, motions of SCR’s will be estimated by the program CABLE3D. CABLE3D is a coupled time domain program that simultaneously predicts the motions of an FPS and its moorings and risers in a specified environment. The program has been well verified with experimental data for TLP’s and spars. A simplistic seafloor boundary condition is now used in CABLE3D. The seafloor stiffness is currently assumed to be linear, and damping and uplift forces are neglected. Some relatively straightforward improvements will be made to CABLE3D to more accurately estimate the loads on the SCR near the seafloor. As the study progresses, learnings regarding the seafloor boundary conditions will be also incorporated in CABLE3D (see below).
SCR and FPS Models - Several models of SCR’s will be developed to bracket a range of diameters (say 18 to 30 inches) and waterdepths (say 3000 to 10,000 ft). Several types of FPS’s (say a TLP, spar, and a semisubmersible) will be analyzed to capture the different motion characteristics that can affect SCR’s.
SCR Motions at the Seafloor - These models will be used to predict the range of SCR motions (amplitudes, velocities, and frequencies) for different Gulf of Mexico environments (hurricanes, loop currents, and normal waves) and FPS/SCR combinations. These SCR motions will be used in the analyses of SCR Bending Stresses described below.
CABLE3D Seafloor Boundary Condition - As the study progresses, learnings regarding the seafloor boundary conditions will be incorporated in CABLE3D to improve the predicted riser motions at the seafloor, which can in turn lead iteratively to further improvements in the seafloor boundary conditions.Parametric Studies of Bending Stresses The magnitude of bending stresses in the riser pipe will be controlled by the combined effects of seafloor stiffness, riser motion characteristics, and touchdown zone characteristics. Parametric studies relating these factors to bending stresses in the riser pipe will be performed using the pipe-spring support model discussed earlier (Figure 1a). The study will consider the following factors:
P-d curve characteristics The studies described in the previous section will provide broad limits on the range of P-d curves associated with various conditions of seafloor stiffness that should be considered in the parametric study. In addition, they will permit modeling of riser-soil interactions for variable conditions of pipe embedment and velocity along the length of the touchdown zone.
Amplitude of riser motions High amplitude riser motions will tend to be associated with a lower overall seafloor stiffness due to the effect yielding of the soil. However, high velocity riser motions will result in increased soil resistance due to strain rate effects. Numerical studies will be conducted to assess the net effect of these counteracting trends on bending stresses in the pipe.
Touchdown zone geometry The geometry of the touchdown zone may be characterized in terms of its length, angle of inclination of the riser pipe at the touchdown point, and maximum trench depth. A predictive model for characterizing this geometry is a formidable task and will not be attempted at this stage. Instead, observational data from existing installations will be used to establish ranges of these variables that should be considered in the parametric study.
Parametric Model for SCR Bending Stresses - The above studies will be used to develop a preliminary parametric model for SCR bending stresses. That model will be useful in providing guidance as to the relative importance of various parameters and estimating the range of bending stresses for various design scenarios.
Assess Needs for Phase 3 - The results of Phase 2 will be used to determine the needs for model refinements, extension to 3D modeling, and/or experimental data to further calibrate and validate the SCR models and recommended boundary conditions.
Deliverables for Phase 2: Interim status reports and a Final Report on Phase 2.
PROJECT PLAN FOR PHASE 3 (2006-2007):
Anticipated Scope of Work & Results - The parametric model developed in Phase 1 will be compared to available laboratory experiments and field data, and calibrated as needed to improve agreement with the data. Additional experiments may need to be undertaken to improve and/or validate the models developed in Phase 2. The model may be also need to be refined or extended to 3D to more accurately predict SCR behavior in the touchdown area. Recommended boundary conditions (stiffness and damping) for use with riser analysis programs such as CABLE3D or other commercially available riser analysis programs in use by the industry will be recommended.
PRINCIPAL INVESTIGATORS AND OTHERS INVOLVED IN THE PROJECT:
PI’s: C.P. Aubeny, G. Biscontin, Jun Zhang, and J.D. Murff
Others: Graduate Students
Date: June 2006Project Title: SCR Touchdown Project
MMS Project: 510 TO Number: 35988
PI: Charles Aubeny, Giovanna Biscontin, Don Murff, Jun Zhang
COTR: Mik Else
Estimated Completion Date: June 30, 2006
Project Description:
The introduction of compliant floating systems for hydrocarbon production has led to the development of new designs for the riser pipes, with the catenary steel compliant riser (SCR) often being the system of choice. Fatigue stresses are critical to SCR performance, and the touchdown zone (TDZ) where the SCR contacts the seabed is often the critical location for fatigue. Analyses show fatigue damage to be sensitive to seafloor stiffness, which cannot be reliably estimated at present. The project objectives are to (1) improve the current state of understanding of the basic mechanisms affecting the seafloor stiffness in the TDZ, (2) develop a model for quantitative estimates of seafloor stiffness, (3) relate seafloor stiffness conditions to SCR bending stresses, and (4) provide a seafloor boundary condition for a steel catenary riser dynamically interacting with a floating structure.Progress:
The geotechnical component of the project comprises two parallel research efforts. The first is the development of a MATLAB finite difference riser-seafloor interaction program. This program models the riser as an elastic beam on non-linear spring supports. Time histories of displacement and rotation are prescribed near the touchdown point. The interaction program produces complete estimates of riser shear, moment, and deflections along the entire length of the touchdown zone. The second geotechnical task is the development of P-y curves to describe the spring supports. Full description of the P-y curves involves definition of riser-seafloor force-displacement relations for four conditions: (1) downward riser motions of virgin penetration of the riser into the seafloor, (2) upward riser motions under conditions of full contact between riser and seafloor, (3) upward riser motions under conditions of breakaway between the riser and seafloor, and (4) downward riser motions under conditions of re-establishment of contact between riser and seafloor. P-y functions describing these relationships have been developed based on force-displacement relationships observed from laboratory model tests and finite element simulations. The formation of a trench within the riser touchdown zone will influence the P-y relationships. The finite element simulations consider the effect of trench width and depth.
COUPLE and CABLE3D, both time-domain programs that simultaneously predict the motions of a FPS and its moorings and risers in a specified ocean environment, have been used to investigate the motions of two SCRs near their TDZ. The SCRs are attached to a truss spar, Horn Mountain, deployed in the Gulf of Mexico in 1,650 m of water. It is found that the motions of the SCRs near the TDZ are dominated by the heave of the truss spar. Although the slow-drifting surge or sway of the spar is much greater in amplitude than its heave, their effects on the motions of the SCRs near the TDZ are relatively insignificant. Predicted motions of the SCRs near their TDZ were provided to the study of the soil resistance of SCRs in the TDZ. The finding that the heave of a spar dominates the motion of a SCR near its TDZ may be applicable to other types of floaters, such as TLP, Semi and FPSO, as long as they are in relatively deep water. Therefore, the magnitude of heave of various Semi, TLP and FPSO in different sea states will be studied and used for further dynamic simulation of the SCRs. The current version of CABLE3D assumes linear soil stiffness and hence is being modified based on the P-y curve provided by the other part of this research project mentioned above.
Reports & Publications:
1. Biscontin, G. & Aubeny, C.P. “Seafloor stiffness model for catenary riser touchdown zone,” in preparation for the International Workshop on Constitutive Modeling – Development, Implementation, Evaluation, and Application, organized by The Hong Kong Polytechnic University and the International Association of Computer Method and Advances in Geomechanics, Hong Kong, China.2. Biscontin, G. & Aubeny, C.P. “Analytical framework for steel catenary riser interaction with the seafloor,” in preparation for submission to the International Journal of Geomechanics.
3. Theckum, B., Numerical Simulatiton of the Truss Sar ‘Horn Mountain’, MS Thesis, Texas A&M University, College Station, Texas, 2006.
4. You, J.H. Numerical Model for Steel Catenary Riser on Seafloor Support, Master’s Thesis, Texas A&M University, College Station, Texas, 2005.
Date: December, 2005Project Title: SCR Touchdown Project
MMS Project: 510 TO Number: 35988
PI: Charles Aubeny, Giovanna Biscontin, Don Murff, Jun Zhang
COTR: Mik Else
Estimated Completion Date: December, 2006
Project Description: The introduction of compliant floating systems for hydrocarbon production has led to the development of new designs for the riser pipes, with the catenary steel compliant riser (SCR) often being system of choice. Fatigue stresses are critical to SCR performance, and the touchdown zone (TDZ) where the SCR contacts the seabed is often the critical location for fatigue. Analyses show fatigue damage to be sensitive to seafloor stiffness, which cannot be reliably estimated at present. The project objectives are to (1) improve the current state of understanding of the basic mechanisms affecting the seafloor stiffness in the TDZ, (2) develop a model for quantitative estimates of seafloor stiffness, and (3) relate seafloor stiffness conditions to SCR bending stresses.
Progress: The first component of the study involves modeling of the soil resistance in the TDZ as a series of soil springs, an approach analogous to P-y analyses of laterally loaded piles. The effects of various types of P-y curves on SCR bending stresses are investigated in this series of analyses. A master’s thesis presenting preliminary work in this area is scheduled for publication in December 2005. The second series of analyses involves two-dimensional finite element studies to develop P-y curves for various conditions of trench geometry, backfilling, and soil softening under cyclic loading.
The portion of the project studying the motions of SCRs near the TDZ was started in September 2005. Firstly, CABLE3D, a coupled time domain program that simultaneously predicts the motions of an FPS and its moorings and risers in a specified ocean environment, is used to investigate the motions of two SCRs near their TDZ. The FPS in this example is a truss spar in 1,650 m of water in a Gulf of Mexico environment. Predicted motions will be applied to the study of the soil resistance in TDZ. Secondly, the linear stiffness model in the current version of CABLE3D is being upgraded to model non-linear soil stiffness using a cubic polynomial. These two near-term objectives are expected to be accomplished before the May of 2006.
Reports & Publications:
You, J.H. Numerical Model for Steel Catenary Riser on Seafloor Support, Master’s Thesis, Texas A&M University, College Station, Texas, 2005.
Date: May, 2005
Project Title: Seafloor Interaction with Steel Catenary Risers
MMS Project: 510 TO Number: 35988
PI: Charles Aubeny, Giovanna Biscontin, Don Murff
COTR: Mik Else
Estimated Completion Date: December, 2006
Project Description: As hydrocarbon production progresses into deep and ultra-deep waters, conventional gravity systems are being replaced by compliant systems comprised of large floating systems attached to the seafloor by vertical tethers or mooring lines. The introduction of these compliant floating systems has led to the development of new designs for the riser pipes, with the catenary steel compliant riser (SCR) often being system of choice. Fatigue stresses associated with extreme storms, vessel movements, and vortex-induced vibrations are critical to SCR performance, and the zone at which the SCR contacts the seabed, the touchdown zone, usually proves to be a spot where bending stresses are largest and therefore a critical location for fatigue. Analyses typically show fatigue damage to be quite sensitive to seafloor stiffness, which at present cannot be estimated with a great deal of reliability. The objectives of this project are to (1) improve the current state of understanding regarding the basic mechanisms affecting the seafloor stiffness within the SCR touchdown zone, (2) develop a means for providing quantitative estimates of seafloor stiffness and its variation over the life of the SCR, (3) identify how various seafloor stiffness conditions affect SCR bending stresses, and (4) incorporate the seafloor stiffness model as a boundary condition in a general model for analysis of riser motions and stresses.
Progress: An analytical framework has been established for this project, and preliminary studies and analyses have been completed to provide an overall scope and plan for the further research and more detailed analyses to be completed in this project. The analytical framework considers the riser-seafloor interaction problem in terms of a pipe resting on a bed of springs (Figure 1), the stiffness characteristics of which are described by non-linear load-deflection (P-d) curves. The? P-d curves will be non-linear and non-uniform along the length of the touchdown zone.
The research approach involves parallel studies to (1) develop P-d curves, and (2) assess the effects of seafloor-riser interactions on bending stresses within the riser pipe.
P-d Curves Factors that can affect the characteristics of P-d curves include properties of the intact soil, the geometry of the trench formed by the riser pipe, backfilling of the trench with remolded soil, soil scour due to currents induced by riser motions, separation of the riser pipe from the soil, disturbance of the soil due to load repetitions, and strain rate effects (see Figure 2).
An immediate focus of this research is to identify those factors that exert the greatest influence on P-d curves and ultimately bending stresses in the riser. Here we show preliminary results of an investigation of trench geometry. Figures 3 compares P-d curves for a narrow and wide trench as predicted by 2D FE analyses. For the narrow trench (W/D = 1), the family of P-d curves shown in Figure 3a show a fairly high sensitivity to trench depth due to the confining effect of the soil. However, analyses for more realistic conditions involving enlarged trench widths (e.g., W/D = 2 in Figure 3b) show a dramatically reduced sensitivity to trench depth. In general, analyses for W/D > 2 indicate a general insensitivity of the P-d curves to trench geometry. If the conclusion holds, it can prove quite useful.
Bending Stresses The P-d curves discussed above will be incorporated into a general framework for evaluating the significance of seafloor-riser interactions on bending stresses in the pipe. The overall framework will consider the factors outlined in Table 1.
Table 1 Parameters Considered in Assessing Bending Stresses in Riser Pipe
Parameter Relevance to SCR Bending Stresses at Touchdown Soil to Pipe Stiffness Ratio elastic stiffness of the soil mass to the flexural stiffness of the pipe Soil Yielding pipe displacement relative to displacement at which significant soil yielding occurs Strength Degradation strength loss in the soil due to repeated cyclic motions of the riser pipe Touchdown Zone Geometry length of touchdown zone relative to the maximum depth of embedment of the riser Strain Rate Effects soil stiffness increases due to increasing strain rates
Investigating the influence of the factors outlined in Table 1 will involve extensive parametric studies. Figure 4 presents preliminary results from such a parametric study for a specific P-d curve. The lower and upper pairs of curves correspond to intact soil undrained shear strengths of 50 lb/ft2 and 500 lb/ft2. The analyses indicate that the effect of touchdown zone geometry, LTDZ/hmax, is relatively small. However, the ratio of soil to pipe stiffness is clearly critical, as is the magnitude of the riser motions near the touchdown point. The non-linear stiffness of the seafloor depends on the magnitude of riser motions, while the stiffness and energy dissipation (damping) characteristics of the seafloor can influence riser motions.
This example illustrates that the SCR-seafloor interaction constitutes a coupled problem, in which the seafloor stiffness and damping characteristics influence the riser motions near the touchdown zone and vice-versa. The 2005-06 research project will therefore include analysis of riser motions in addition to the continuing geotechnical studies.
Date: December, 2004
Project Title: SCR Touchdown Project
MMS Project: 510 TO Number: 35988
PI: Charles Aubeny, Giovanna Biscontin, Don Murff
COTR: Mik Else
Estimated Completion Date: December, 2006
Project Description: As hydrocarbon production progresses into deep and ultra-deep waters, conventional gravity systems are being replaced by compliant systems comprised of large floating systems attached to the seafloor by vertical tethers or mooring lines. The introduction of these compliant floating systems has led to the development of new designs for the risers, with the catenary steel compliant riser (SCR) often being system of choice. Fatigue stresses associated with extreme storms, vessel movements, and vortex-induced vibrations are critical to SCR performance, and the zone at which the SCR contacts the seabed, the touchdown zone (TDZ), usually proves to be a spot where bending stresses are largest and therefore a critical location for fatigue. Analyses typically show fatigue damage to be quite sensitive to seafloor stiffness, which at present cannot be estimated with a great deal of reliability. The objectives of this project are to (1) improve the current state of understanding regarding the basic mechanisms affecting the seafloor stiffness at the SCR touchdown point, (2) develop a means for providing quantitative estimates of seafloor stiffness and its variation over the life of the project, and (3) identify how various seafloor stiffness conditions affect SCR bending stresses.Progress: Two series of analytical studies are being performed to achieve the above objectives. The first series of analyses involves modeling of the soil resistance in the TDZ as a series of soil springs, an approach analogous to P-y analyses of laterally loaded piles. The effects of various types of P-y curves on SCR bending stresses are investigated in this series of analyses. The second series of analyses involves the development of P-y curves for various conditions of trench depth, trench width, strain rate dependence of soil properties, soil stiffness degradation under cyclic loading, and pore water pressure redistribution. Equivalent P-y curves are being developed based on two-dimensional finite element analyses. The initial studies are utilizing a simple elastic-perfectly plastic soil constitutive model. Subsequent refinements will be added to this model to simulate various complexities of soil behavior such as stiffness degradation under cyclic loading.
Reports & Publications: The studies described above are in their initial phases and no publications have been completed to date.
