Deepwater Mooring Systems: Concepts, Design, Analysis, and Materials

Laboratory experiments are being conducted to study the behavior of suction caissons used for deep offshore moorings. Results from nine caisson pullout tests were used to quantify the components of axial capacity. Tests with a 100-mm diameter caisson prototype, installed using dead weight or suction, were performed in a 1.1-m thick deposit of kaolinite. Large tanks of normally consolidated clay, selected to simulate common seafloor conditions, were prepared by consolidating a kaolinite slurry. Axial pullout tests were carried out on caissons inserted using dead weight only or dead weight plus suction pressure, on caissons pulled with a vented or sealed top cap, and with rapid (undrained) versus slow (drained) pullout. Measured axial capacities are interpreted in terms of the weight of extracted soil, side resistance on the caisson walls, and reverse end bearing capacity at the tip. For rapid pullout with a sealed top cap, the test results indicate an external side resistance factor (α) of 0.5 to 0.8 and a reverse end bearing factor (N'_c) of 13 to 21.

Three centrifuge model tests were performed in normally consolidated Speswhite Kaolin to study the penetration of suction anchors in soft clay. The suction anchors could be penetrated by means of underpressure to a depth of about 12.4 to a little more than 14.5 times the diameter. When the anchors were penetrated by underpressure, all clay displaced by the skirt moved into the anchor. At a penetration depth of about half the maximum penetration depth, the volume of the soil heave inside the anchor actually increased more than the volume of the displaced clay. When a material coefficient of 1.5 against plug failure was mobilized, more than the clay displaced by the skirts had accumulated inside the anchor. The penetration resistance increased by 42% and 26% during rest periods of 4.5 and 0.8 days prototype time, respectively. Some uncertainty in the shear strength of the clay beds gave some uncertainty with respect to the interpretation, but the observed behaviour generally confirmed the theoretical analyses.

Key aspects of the suction anchor pile design for the Na Kika Floating Development System are presented. Site characterization of the foundation soils, determination of the required pile size and structural design are emphasized. Soil parameters for suction anchor pile design are developed using an integrated approach, incorporating shallow geophysical data, geotechnical data obtained from soil borings and anchor installation records of mobile offshore drilling units. The locations of soil borings were selected based on the principal geology of the development area, and not on specific anchor locations. Spatial variation of soil properties is evaluated in the context of the geologic model and available data. A simple procedure is demonstrated to estimate distributions of overburden pressure and undrained shear strength at specific anchor locations. The required pile size is based on a target probability of failure of the overall mooring system. With probabilistic models of load and resistance, event tree analyses are used to evaluate design preferences with respect to failure modes and estimate the annual probability of failure. The design also accounts for spatial soil variability, shear strength anisotropy and cyclic load effects. Most of the pile cylinder experiences negligible stress when subjected to in-service loads. The section of the pile local to the mooring line connection, however, is limiting in terms of structural strength and fatigue. Ultimate strength analysis of the pile indicates an ultimate (structural) strength greater than the pile capacity (with respect to soil failure) and the chain break load.

The pertinent installation considerations and field performance of suction anchor piles for the Na Kika Floating Development System are presented. Several aspects of the pile structure are determined directly by installation considerations, rather than in-service loads. Critical buckling pressure during installation governs the nominal wall thickness of the pile cylinder, while maximum pile recovery pressure controls the pile top lid and associated stiffeners. A pile shoe is added for a robust structure during handling and pile penetration. The predicted underpressures during pile installation are based on penetration analyses that account for pile geometry, including structural stiffeners, and expected soil resistance. Actual measurements of self-weight penetrations and applied underpressures concur with design predictions at each of the four anchor groups. Installation behavior of the piles was generally as anticipated, but an occasional tendency for the pile to rotate significantly during the installation process was unexpected. Implications of the pile installation performance for the Na Kika mooring system and future designs of suction anchor piles are also discussed.

Upper bound plastic limit analyses (PLA) can provide a useful framework for estimating the load capacity of suction caisson anchors in purely cohesive soils. Since arbitrary assumptions regarding the soil stress state are not required in the PLA formulation, it may be used with greater consistency compared to other simplified approaches such as limit equilibrium methods. While PLA methods do not attempt to include all of the complexities of anchor behavior, they can provide a relatively simple framework for visualizing anchor kinematics leading to an understanding of the relative importance of various parameters on suction anchor load capacity. The most rigorous PLA formulations involve postulating a three-dimensional anchor-soil failure mechanism and deriving expressions for internal energy dissipation throughout the mechanism. This approach can involve extensive numerical integrations and a relatively complex scheme for optimizing the failure mechanism to obtain a least upper bound collapse load. Considerable simplification is possible if the problem is formulated in terms of ultimate unit resistances (lateral, axial, and their interaction) that can be exerted by the soil on the caisson. In this case the caisson failure mechanism can be characterized in terms of one or two optimization variables. Simple expressions for the ultimate unit resistances acting on the caisson can be obtained from several sources including rigorous PLA solutions, finite element techniques, or experimental measurements. General expressions are possible by limiting consideration to common, idealized strength profiles such as uniform or constant gradient. Such simplified formulations are particularly valuable for providing an analysis tool accessible to practicing engineers. Suction caisson anchors can be subjected to a variety of load orientations including nearly vertical uplift forces imposed by the vertical tendons of tension leg platforms, horizontal loads imposed by catenary mooring systems, and inclined loads imposed by taut moorings. Recently, PLA methods have been applied to the analysis of suction caissons subjected to this range of loading conditions. This paper reviews the formulation of these analyses and summarizes the most significant findings.

Computational tools are being developed for predicting the response of synthetic-fiber ropes under both monotonic and cyclic loads. A scalar quantity, termed the damage index, is included in the model for evaluating the degradation of rope properties with load. The damage index is used to modify virgin rope behavior and simulate the behavior of ropes that experience damage throughout their loading history. In this paper, static tension tests are used to obtain the evolution of the damage index as a function of the strain history. This evolution is computed based on the stiffness ratio between experimental data and simulation of rope response without any source of damage. An exponential damage index evolution law is used for computational purposes. Numerical simulations of damaged rope behavior compare quite well with experimental data.

Polyester ropes, are used in a taut mooring system, especially in the case of a floating offshore structure deployed in deep water. Previous studies were conducted concerning mechanical properties of polyester mooring lines, providing a solid foundation for dynamic simulation of polyester mooring lines. In addition to light wet-weight and relatively large elongation of polyester ropes, experimental studies indicated that the modulus of a polyester rope depends on the static and dynamic tension. In this study, we developed a method to use an empirical formula of previous studies to model the modulus of a polyester rope based on its mean and dynamic loads and then simulate dynamic interaction between a classical SPAR and a hybrid polyester mooring system in the time domain. Through the simulation, it is also found resonant vibrations in a polyester rope may be generated and eventually lead to the failure in numerical simulation. To alleviate the building up in resonant tension in numerical simulation, proper elastic damping in a polyester rope should be introduced.

The Mad Dog Project will use a polyester mooring system on the drilling and production truss spar. This will be the first use of a permanent polyester mooring system on a Floating Production System (FPS) outside of Brazil and the first time polyester has been used on a spar. As such, there were many challenges, which include regulatory approval; designing a mooring system which is dominated by current loadings; largest polyester rope break load ever required; rope design qualification and testing; quality control and assurance; and inspection, maintenance, repair and retirement (IMRR) of such a mooring system. This paper will focus on the design issues, rope design, manufacturing, qualification and testing, and the IMRR plan that was developed to provide the assurance that the polyester mooring system could be safely operated over the 20-year field life.

The polyester lines for mooring of deepwater platform in a taut leg configuration are a reality offshore Brazil, holding in place semi-submersible platforms. With that, the need to simulate these lines in model tests became a necessity. Although the non-linear behavior is clear, depending on the type of cycling, the polyester rope responds in ways that may be considered linear as a steel wire rope. It is not clear so far how this is happening but anyway the behavior is clear since it repeats itself after same load history. For early model tests that have to be performed as usual urgently, the experts decided to use steel springs with a linear restoring capability. When the dominant restoring coefficients changes, then the spring have to be replaced by new ones and, probably, the tests repeated. The use of equivalent springs seemed the proper way. However, with the help of fundamental investigation on the similarity laws, the present work shows that the use of very thin polyester lines in model scaling is feasible and will indeed allow a closer physical representation. The proposal is to use polyester lines with length distortion in order to keep the similar restoring capability. By avoid using springs, but using the same polyester material as in full scale, the same non-linear behavior is present during the tests. As shown, even the response to random excitation due to random waves is closer represented.

The theory of three-dimensional hydroelasticity for free flexible floating bodies and the theory of mooring system analysis are integrated into a complete theoretical framework for analyzing the dynamical response of the mooring system of a flexible floating body. The effect of the elastic deformations of the floating body on the mooring system analysis is investigated. The mathematical model based on the perturbation method for the dynamic tension response analysis of the mooring system is presented. Three-dimensional hydroelasticity theory and Goodman-Lance method are used to solve the hydroelastic response of the floating body and dynamic tension response of the mooring line respectively. Their coupling relationship are also given and used to solve motion equations of the mooring lines. The theory presented in the paper is illustrated by a numerical example. And the results show that the elasticity of the very large floating structure has considerable effect on the dynamic responses of the mooring system.

Fully coupled global analysis of Floating Production Systems, including the vessel, the mooring system and the riser system is described. Design of the system can be a daunting task, involving more than 1,000 load cases for global analysis. The primary driver for the mooring system and for the riser system is motion of the vessel. Vessel motions are driven by environmental forces, but are restrained by forces from the mooring and riser systems. Numerical models and procedures that provide accurate and efficient global modeling of the Floating Production System are presented. Both Time Domain and Frequency Domain procedures are included. The accuracy and efficiency of the procedures are illustrated in an example: a large semi with 16 mooring lines and 20 risers. The procedures provide the accuracy and efficiency for use of fully coupled analysis in design of Floating Production Systems from concept selection to final design, installation and operation.

A numerical code (COUPLE) has been developed for computing 6 Degree-Of-Freedom (DOF) motions of a moored floating structure dynamically interacting with its mooring/riser/tendon system. The computation of potential wave load on a moored structure can be conducted based on a diffraction wave theory, e.g. WAMIT, or the Morison Equation based upon a slender body assumption. In-line drag and transverse force caused by currents and wave induced particle velocity are computed using a modified Morison Equation. Wave kinematics up to the free surface used in the Morison Equation is computed using nonlinear deterministic Hybrid Wave Models, and is accurate up to second order in wave steepness. Wind forces are computed based on empirical formula suggested by the related API recommendation.

In global motions and mooring analysis for a Truss Spar, the mooring system is usually modeled either quasi-statically or coupled dynamically, while the modeling of riser system is more complicated. This is especially true with large numbers of risers attached to a Spar located in deepwater. In this paper the motions and mooring responses of a Truss Spar under a typical 100-year extreme sea state in three different water depths ranging from 2000ft to 6000ft are analyzed. The mooring system is modeled either quasi-statically or coupled dynamically. For the riser system, two types of risers, Top Tension Risers (TTRs) and Steel Catenary Risers (SCRs) with three different models are investigated: 1) fully coupled dynamics with a finite element model, 2) top tension provided from catenary equation quasi-statically, and 3) a constant force. The ABS in-house Coupled Analysis Program (CAP), which is capable to apply different models for simulating the mooring and riser systems, has been used for this study. The differences in, system stiffness, added-mass, added-damping, and Spar motions as well as mooring loads have been derived for various mooring and riser models in different water depths. The results of these differences are presented and discussed in this paper. The findings from this study will be useful for the design and analysis of vessel motions and mooring responses for deepwater floating installations.

This paper aims to investigate basic interaction characteristics of side-by-side moored vessels both numerically and experimentally. A higher-order boundary element method(HOBEM) combined with generalized mode approach is applied to analysis of motion and drift force of side by side moored multiple bodies(LNG FPSO, LGC and shuttle tankers). Model tests were carried out for the same floating bodies investigated in the numerical study in regular and irregular waves. Global and local motion responses and drift forces of three vessels are compared with those of calculations. Discussions is highlighted on applicability of numerical method to prediction of sophisticated multi-body interaction problem of which motion behavior is very important to analysis of mooring dynamics of deep sea floating bodies.

The computational analysis of the multi-point radiation moored system, which composed of mooring lines with buoys for a semi-submerged drilling platform working in deep water, is completed by use of the quasi-static method, dynamic method and matching approach for platform and mooring line-buoy systems in this paper. The motion equations for floating structures are solved by the Green function method with numerical panel approach and the mooring line-buoy kinetics and kinematics problem is formulated as a combined nonlinear initial-value and two-point boundary-value problem. Then, the motion equations both for floating structure and for mooring line-cable can be solved by use of the matching approach. Thus, the motion of the floating structure and buoys and the tension of the mooring line are determined.

The oil industry is concentrating its efforts and activities on developing the Turret mooring system in deep waters. These waters range from 1,000m – 3,000m and include areas the Gulf of Mexico, Brazil, West of the Africa and North Atlantic/Norwegian Sea. The mooring stiffness of vessel which is moored in the deep sea is softer than the one in shallow water. This causes large non-linear dynamic effects on the moored vessel in the ultimate sea state. The accurate prediction of low frequency motions of FPSOs is essential for safe design of mooring and riser systems. For ultra deep water, the modeling of full-depth mooring system is very difficult since no tank facility is large enough to perform complete system tests with reasonable depth for model scale. In this study, a combination of model test with truncated set-up and computer simulations — the hybrid approach — is suggested. The damping terms on FPSOs, such as wave drift damping, are very important dynamics of moored vessels. The amount of slow drift damping of a FPSO in the still water and regular waves is measured through free decaying tests. By the source and dipole distribution method, the first-order and second-order wave exciting forces are calculated. Quasi computed motions of the low frequency surge of a turret moored FPSO are simulated, using added masses and damping coefficients measured from model tests. The simulated motions are compared with test results.

A vessel/mooring/riser coupled dynamic analysis program in time domain is developed for the global motion simulation of a turret-moored, tanker based FPSO designed for 6000-ft water depth. The vessel global motions and mooring tension are simulated for the non-parallel wind-wave-current 100-year hurricane condition in the Gulf of Mexico. The wind and current forces and moments are estimated from the OCIMF empirical data base for the given loading condition. The numerical results are compared with the OTRC 1:60 model-testing results with truncated mooring system. The system's stiffness and line tension as well as natural periods and damping obtained from the OTRC measurement are checked through numerically simulated static-offset and free-decay tests. The global vessel motion simulations in the hurricane condition were conducted by varying lateral and longitudinal hull drag coefficients, different mooring and riser set up, and wind-exposed areas to better understand the sensitivity of the FPSO responses against empirical parameters. It is particularly stressed that the dynamic mooring tension can be greatly underestimated when truncated mooring system is used.

Model tests of a DICAS moored FPSO have been carried out through 1:90 scaled experiments. The vessel motions and the mooring line tensions are investigated. The full-scale depth is 1240m. The model tests are performed with a mooring and riser system model truncated at 450m water depth. Extreme metocean conditions for Brazilian waters are simulated, including waves, current and wind in collinear configuration. A time-domain global hydrodynamic analysis of the truncated system is also carried out. Initially, wave loads on the hull are modelled by use of linear diffraction analysis (WAMIT), including mean loads and wave-frequency excitation. Low frequency excitation is included by use of Newman's approximation. In addition, viscous drag loads on the hull due to relative motions in waves and current and wind loads are included through quadratic current- and wind coefficients. The vessel motions are dynamically coupled in the time domain to the top-end mooring line and riser tensions, through detailed FEM (Finite-Element-Modelling) assuming slender-body hydrodynamics of this part of the system. The RIFLEX-C software is used. Irregular wave records from the experiments are used as input to the simulations, making direct comparison of response time series and spectra possible. Results from numerical simulations are compared with model test results. Based on the comparisons, hydrodynamic parameters of the numerical model are adjusted to match the measurements. In particular, drift force coefficients and thus slow-drift excitation, as well as damping coefficients, are empirically modified. Wind and current coefficients are calibrated by comparisons to experiments where the FPSO is exposed to pure wind and current conditions, respectively. Resulting simulations using the calibrated numerical model agree well with experimental results, and represent significant enhancements relative to the initial model. Using the calibrated hydrodynamic vessel coefficients, the numerical model can be extrapolated to full depth. Accordingly, numerical simulation of the full depth system can be performed. In the present paper, the focus is on the model tests and the calibration of the numerical model. The extrapolation phase, which is more straightforward, is not emphasized here.

A study has been performed to find mooring line safety factors that would maintain the present safety level of the Norwegian offshore regulations if the present possibility of optimising the system by adjusting the mooring line pre-tension during operation were taken out. This may become necessary if the Norwegian authorities want to harmonise their regulations with the up-coming ISO 19904 standard for floating offshore structures. Design analyses without tension optimisation generally leads to higher tension than with optimisation. However, simply disallowing line adjustment in design would tend to increase the required strength of the resulting mooring line designs, and thereby raise the safety level. Thus, in order to maintain the safety level if line tension optimisation is excluded, the present safety factors should be reduced. Calculations have been performed for permanent installations and mobile units, and in each case a "new" safety factor is proposed.

Simultaneous optimisation of riser- and mooring systems for floating production units has been made possible by combining efficient design analysis tools for mooring and riser systems with a state-of-the-art nonlinear optimization code. This can contribute to improving the conceptual design work of anchor and riser systems. The fatigue load may become more critical than the extreme load in some cases, depending upon environment conditions, and floater- and mooring system properties. Thus, the minimum required service life should be included as a design constraint in the optimisation problem formulation. The feasibility of doing this has been demonstrated by implementation in the computer program MOOROPT. An efficient frequency domain analysis of vessel motions and of mooring line dynamics allows analysis of a large number of cases for each gradient evaluation in the optimum search scheme. The sensitivity to the number of seastates analysed is investigated, and it is shown that condensing the `scatter diagram' from 1400 to about 40 – 60 cases, covering 8 direction sectors of environmental conditions may be acceptable for representing the fatigue life.

This paper describes the use of coupled analysis for integrated design of riser and mooring systems. Coupled analysis, which includes the mooring lines, risers and vessel in a single model, is required to accurately account for interaction effects. For example, lateral loads on the vessel from the risers, which can be of the same order as extreme environmental forces, must be resisted by the mooring system. Stiffness and drag-induced damping effects due to the risers can also be significant. Accounting for the interaction effects should lead to more efficient designs. Integrated riser and mooring design requires a substantial amount of analysis early in the design cycle. Consequently, the analysis method must be both accurate and efficient. We describe a fully coupled, frequency domain analysis method that provides the required accuracy and efficiency. Integrated design of mooring and risers is illustrated with an example: an FPSO moored in deep water offshore W. Africa with 16 mooring lines and 20 steel catenary risers (SCRs) installed in stages. An important aspect of the example is that it includes consideration of riser fatigue life in the integrated mooring and riser design. This is accomplished by calculating SCR fatigue life as a function of mooring pretension. This illustrates how a mooring may be designed based on SCR performance rather than on an arbitrary vessel offset limit.

There is an increasing amount of laboratory and field data being gathered on the behavior of marine risers with the objective of developing a better understanding their dynamic behavior for design purposes. The focus of this discussion is the effective utilization of the reverse system identification technique for offshore applications and in particular the consideration of relevant practical considerations for its application to marine risers. Some of the critical issues include consideration of the importance the frequency dependence of key model parameters, the use of fully correlated data, the form of the nonlinear dependencies, and the need for a variety of input and response measurements.

Large motion responses of offshore platforms in harsh environment in deep water are of great concern for the operation and design of the platforms. Thruster-assisted position mooring systems can be applied to avoid excessive motions of the platforms. Fully coupled dynamic analysis of a thruster-assisted turret-moored Floating, Production, Storage, and Offloading unit was performed. Kalman filtering, PD control, and thrust allocation were linked with the hull/mooring/riser time-domain coupled dynamic analysis program called WINPOST-FPSO. The numerical example illustrates general performance of a thruster-assisted turret-moored FPSO and shows that the mooring/riser tensions are reduced.

Although Morison equation is often applied for predicting hydrodynamic force of marine structure, it may give poor predictions when non-linear behavior is severe or random wave is encountered. The problem can be solved by system identification technique. From the dynamic system point of view, the velocity of heave motion is regarded as a system's input and hydrodynamic force the output. In this paper, NARMAX model theory is used to identify the hydrodynamic system of heave damping plate. Linear and non-linear model are obtained and the comparison between the results of predicted and measured showed that they can predict the hydrodynamic response better than Morison equation. The measured data from forced oscillation tests used for identification are random records with given spectrum. The forced oscillation tests are also introduced in this paper. Numerical computations are also performed using vortex tracking method presented by Faltinsen and Pettersen. In computations, three dimensional heave plate is simplified into two dimensional vertical plate. We obtain some interesting results.