DESIGN CONSIDERATIONS FOR SELECTION OF FLEXIBLE RISER CONFIGURATION

PD-Vol. 42, Offshore and Arctic Operations - 1992 ASME 1992 DESIGN CONSIDERATIONS FOR SELECTION OF FLEXIBLE RISER CONFIGURATION N. Ismail, R. Nielsen, and M. Kanarellis Wellstream Corporation Panama City, Florida ABSTRACT A brief review of recent literature on riser design is presented and a concise description of the design process is given. Effects of hydrodynamical design parameters on marine flexible riser design are then reviewed. Riser dynamic analyses are described which substantiate design criteria for the selection of riser configuration in deep and shallow water. The results of dynamic analyses using computerized numerical models are presented in this paper in graphical form for the selected riser design cases. Output includes envelopes of riser coordinates, axial force and time histories of forces and wave surface profiles. The results obtained highlight the significance of motion spatial gradients for the combined flow of waves, currents and vessel heave motion on the selection optimum of riser configuration to address design requirements. NOMENCLATURE A = flexible pipe cross-sectional area H L T a g h K = water pressure head (h-s) = wave length = wave period = wave amplitude = acceleration of gravity = water depth = wave number L 2π Nomenclature (continued) s p Greek Symbols: ρ σ Subscript: i o = elevation above sea bottom = fluid pressure = fluid density INTRODUCTION = wave frequency 2π L = pipe internal properties = pipe external properties Though flexible pipe as a marine product was introduced to the offshore market in the early seventies, it was not until 1978 that flexible risers were specified and installed in the Enchova field offshore Brazil (Machado, 1980) as part of a floating production system. Since 1980, the use of flexible pipe has spread worldwide and is used in almost every offshore oil development today as witnessed in papers by Mahoney (1986) for North Sea application, Tillinghast (1990) for Gulf of Mexico, Gulf of Suez applications, Tillinghast (1987) and Beynet (1982) for the Far East. This type of dynamic application is typically used for floating production systems for high pressure production risers, export risers, chemical/water/injection lines and gas lift lines. Currently, the main manufacturers of flexible pipes are Coflexip, Wellstream and Furukawa. References and Illustrations at end of paper 1

At the present time, much interest in riser systems is shown by the operators as evidenced by papers by Beynet (1982) and Ashcombe (1990). RISER CONFIGURATION SELECTION Industry practice calls for several types of riser configurations typically used in conjunction with Floating Production/Loading Systems. The standard five configurations generally used are: Free-Hanging Catenary, Lazy-S, Lazy Wave, Steep-S, and Steep Wave. Figure 1-a illustrates these typical types of riser configurations. Figure 1-b illustrates a schematic of a new riser configuration proposed by Wellstream for the Alcorn Linapacan Field Development Project. The motivation for and validity of this new riser configuration is presented in this paper. The dynamic response of a particular riser system is directly related to the environmental loadings due to the combined wave-current field flow and the dynamic boundary conditions of the riser top end at the water surface, coupled with the interaction arising from the structural nonlinear behavior of the riser itself. To illustrate, design parameters impacting the suitability of a particular configuration for a particular water depth, two riser design cases were studied in deep and shallow water. The computer analyses were carried out at Wellstream s engineering offices using computer program FLEXRISER-4 developed by Zentech, UK. The results of the dynamic analyses for these several design cases are the essence of this paper. The output illustrates the critical aspects of wave and current hydrodynamics as well as vessel motion response affecting the selection of the riser configuration. FLEXIBLE RISER ANALYSIS AND DESIGN Flexible pipes and risers are critical components for offshore field developments because they provide the means of transferring fluids, or power, between subsea units and a topside floating platform, or buoys. These risers accommodate floating platform motion and hydrodynamic loading by being flexible. In storm conditions, they undergo large dynamic deflections and must remain in tension throughout their response. They are consequently manufactured to possess high structural axial stiffness and relatively low structural bending stiffness. Their global dynamic behavior can be considered as more mechanical, or force dependent, than structural. In contrast, behavior near the end connectors of a system is governed by local structural stiffness properties. DESIGN CRITERIA Efficient design of flexible riser systems is made possible by using computer-based solution techniques. The design criteria of flexible riser systems is usually based on allowable pipe curvatures and tensions prescribed by the pipe manufacturer, clearances between the riser and other structures, and boundaries during its dynamic response. The allowable curvatures and tensions are based on full-scale test procedures and stress analysis carried out by the manufacturer. These limits ensure the pipe is not over-stressed when responding to dynamic loads and vessel motions. The system is generally designed so the pipe is tensioned throughout its dynamic response cycle. Minimum clearances are also specified to avoid clashing problems between riser and seabed, or riser and vessel, and between the riser or other adjacent risers, cables, or mooring systems. DESIGN PARAMETERS AND PROCEDURES The main problem in designing flexible riser systems is the large number of design parameters. The environmental conditions, vessel or calm buoy motions and riser properties are usually well-defined. The main design parameters are the choice of riser configuration, the length of riser, the system geometry and the sizing of buoyancy modules, subsurface buoy or arch. The choice of riser configuration is usually based on economic criteria, position of the wells, wave and current forces, motion response and excursions of the vessel or surface buoy as well. The design procedure can be described as consisting of three stages. First Stage The first stage in designing a flexible riser system is determining an acceptable system layout. The first stage is based on static analysis. It is normal to carry out a parametric study assessing the effect of changing the design parameters (i.e., system geometry and length) on the static curvature and tension. Based on the results of this parametric study, the design selects a suitable range of system geometries and lengths satisfying the design criteria. The parametric study will also assess the static effects of vessel offset (displacement of the top end) and the current loading in different directions. 2

Second Stage The second stage in the design procedure is performing a dynamic analysis of the system to assess the global dynamic response. A system layout and length is chosen from stage one and a series of dynamic load cases are considered. These load cases combine different wave and current conditions, vessel or surface buoy positions, and riser contents in order to prove an overall assessment of the riser suitability in operational and survival conditions. The corresponding analyses are then carried out and dynamic curvatures, tensions and clearances are checked against the design limits. The majority of riser dynamic analyses packages, including FLEXRISER-4, make use of the concept of effective tension (Sparks, 1983). Sparks addressed the drilling riser case where the riser is essentially restrained. A catenary riser on the other hand turns 900 to meet the sea bed. It is subject to friction and can be subject to compression due to these conditions. This concept accounts for the effects of external and internal hydrostatic pressure acting on the internal and external surfaces of the pipe wall. It is the effective tension which controls the stability of the riser from the point of view of deflection. The relationship between effective tension, T eff, and the true wall tension, T wall, that acts on the pipe wall and contributes to stress in the pipe wall is: T wall = T eff + (pi ± ρjg H i ) A i (ρ 0 g H O )A O.(1) where: T wall =Wall tension to be used for stress calculation in flexible pipe wall. T eff = Effective tension as predicted by the riser analysis computer program. The effective tension is independent from internal and external pressure. Given the effective tension, as predicted by the riser global analysis program, the true wall tension may be simply calculated from the equation (1). Since internal pressure affects the T wall, it is important to carefully note the internal pressure conditions in the pipe under the maximum load cases as well as the limiting operational conditions when pressure in the riser may be released or maintained. Third Stage The third stage in the design procedure is performing detailed static and dynamic analyses of local areas to design particular components. This state is presented in a separate publication (Brown, 1989). Key papers by operators in this regard are Out (1989), Boef (1990) of SIPM and de Oliveira (1985) of Conoco. This third stage of design also includes a question of life expectancy which has recently been addressed by Claydon, et al (1991). All of the stated design aspects are important but the solution to each problem starts with the selection of an optimum configuration which is the subject of the remainder of the paper. HYDRODYNAMIC LOADING ON RISERS The determination of surface wave hydrodynamic loads on the marine systems are based on two major techniques. For large structures, the scattering of incident waves is considered and a diffraction theory is employed. The wave loads on the small members of flexible pipes are determined by applying the Morrison equation. This equation separates the hydrodynamic loads into inertia and drag forces. Considering the pseudo static design approach (deterministic), it has been customary to use irrotational wave models to predict fluid particle kinematics for a certain design wave. Drag and inertia force coefficients are then used to relate the particle kinematics to the hydrodynamic forces expressed by the Morrison equation. Further, it is important to determine appropriate drag and mass coefficient as well as to adopt the appropriate wave theory representing the design wave characteristics. To include currents in wave force calculations, the oil industry has traditionally used the technique of linear super-position. When doing numerical analysis of the hydrodynamic loading and selection of flexible risers configuration, one has to recall the fact the global riser response in a particular water depth is affected by the spatial and temporal distributions of the integral properties of waves (mass, momentum, pressure and energy); therefore, these distributions of water-wave properties significantly dictate the suitability of a particular riser configuration. To illustrate some of these water wave properties, consider a progressive wave moving in the positive horizontal axis (figure 2). For simplicity, considering the small amplitude wave theory, the wave kinetic energy concentration averaged over one wave period at any elevation above the bottom is given by: a 2 ρg 2 k 2 KE(s) = cosh 2ks...(2) 4σ 2 (cosh kh) 2 In order to represent the above relation in dimensionless form, the ratio of the kinetic energy at any elevation s to the mean free 3

surface is given by KE(s) cosh 2ks =..(3) KE(h) cosh 2kh Equation (3) is plotted in figure 3 for the cases of h/l= 0.05 (conventional shallow water limit) and h/l = 0.5 (conventional deep water limit). The figure shows in the case of shallow water waves, the energy concentration is nearly uniform with depth; in the case of deep water waves, the energy is concentrated near the surface. It should be noted that effects due to wavecurrent interaction should always be considered in the design and analysis because it might cause a significant change in the magnitude and distribution of fluid forces. This change of fluid forces could be dramatic in coastal waters, as seen by Ismail (1984), not only due to the strong current shear, but also due to the change in characteristics of shoaling waves. The parameters usually sensitive to wave-current interaction effects include hydrodynamic force coefficients, current velocity profiles, and relative direction of waves and currents. PRESENTATION AND INTERPRETATION OF RESULTS In order to illustrate the potential and the limitations, of adopting specific riser configurations in a particular water depth, the results of riser dynamics analyses for two design cases in deep water and one design case in shallow water are presented. The computer analyses were conducted at Wellstream s engineering office using the FLEXRISER-4 personal computer program developed by Zentech, UK. FREE-HANGING CATENARY RISERS A Free-Hanging catenary configuration was considered for riser systems in two design cases. The first case is in a 600 m water depth and the second in a 350 m water depth. Table I illustrates the main design parameters used as input for the computer analyses. For the first case, figure 4 shows a snapshot of the riser configuration and the distribution of axial force along the length of the riser depicting a small amount of compression near the seabed. In contrast to this case (for the 350 m water depth), figure 5 shows an appreciable amount of compressive force in the riser near the seabed. To maintain the integrity of a flexible riser system, it is imperative that the riser pipe remains in tension throughout its operational life. Because of the development of this excessive amount of compressive force in the pipe, it can be safely concluded a Free-Hanging catenary configuration is unsuitable for this design case in 350 m water depth. A major reason of the development of axial compression forces is that the heave motion associated with the calm buoy is appreciable. To mitigate the effects of this loss of tension in the riser pipe, which could adversely affect the service life of the pipe, one of the following measures may be adopted: Disconnection of the riser under strong wave conditions; Increase the weight of the riser pipe section near the seabed by wrapping with heavy material; Modify the buoy design to reduce the heave motion; Attach a subsurface buoy, near the seabed. EASY-TOUCH CATENARY RISER Each of the above remedies to eliminate compression forces could be viewed as a possible solution depending on the technical and economical constraints of the project in general and the riser system in particular. The latter concept of adding a subsurface buoy (= 60 kn) to the free hanging catenary riser in the 350 m water depth design case to eliminate compression forces was implemented. The results of dynamic analyses demonstrating the success of the concept are shown in figure 6. It is important to highlight this, if the position of the surface buoy is to increase above the seabed, the riser will be approaching the Steep-S riser configuration, compression forces would still develop over the lower catenary portion. Wellstream had applied this modified concept for the catenary type configuration in the riser design for Alcorn s West Linapacan Development Project in the South China Sea. The Easy-Touch Catenary was coined as the name for this modified catenary configuration. The validity of the newly proposed riser configuration was tested under various operating scenarios for the calm buoy and under normal and maximum environmental conditions. STEEP WAVE RISERS Dynamic riser analyses were conducted for a design of a single oil export system in a 40 m water depth site. An extensive analysis was 4

carried out prior to arriving at the proposed steep riser configuration (either 6-inch or 8-inch flexible pipe diameter) resulting in many unacceptable configurations, primarily due to the riser s inability to withstand the severe environmental conditions imposed by extreme wave heights of 15 m for the I year-storm and 20 m for the 100 year-storm. The analyses showed neither Lazy-S nor Free-Hanging systems were feasible. Moreover, the need to eliminate the Steep S configuration, as an alternative, became apparent when the sensitivity of the riser s dynamic response to the intensity of additional buoyancy distribution was determined. Thus, the Steep Wave was left as the only feasible configuration for the riser of this oil export system. The distribution of buoyancy along the Steep Wave riser was determined (figure 7) to prevent any buckling instability and to reduce any excessive angle at the riser base. A snapshot of the riser configuration under the dynamic effects of waves and currents is shown in figure 7 for both cases of far and near field. The corresponding axial tension force along the Steep Wave riser is shown on figure 7. CONCLUSIONS It was evident the most apparent riser configuration does not necessarily provide the appropriate solution for the design cases of flexible marine risers in deep and shallow water considered in this paper. The deep-water cases emphasize a solution based on a simple riser configuration to facilitate modularity and ease of installation and removal either the standard Catenary in the 600 m water depth case or, the Easy-Touch Catenary in the 350 m water depth case. In the shallow-water case, the design is a more complex riser configuration due to the severe environment loads requiring particular design configuration loads. Both cases represent new frontiers for use of flexible pipes as marine nsers. The sensitivity of the riser dynamic response, in particular configuration to environmental data and vessel/floater motion data, warrants a careful review of design basis prior to the dynamic analysis and design of marine risers. The selected configuration then determines many design parameters, among others, design life and bending stiffener and restrictor requirements. REFERENCES American Petroleum Institute, 1987, RPI7B - Recommended Practice for Flexible Pipe, Houston, TX. Ashcombe, G.T., and Kenison, R. C., BP Engineering, 1990, The Problems Associated with NDT of High Pressure Flexible Pipes, Society of Underwater Technology Conference, Aberdeen. Beynet, P. A., and Frase, J. R., 1982, Flexible Riser for a Floating Storage and Offloading System, Proceedings of Offshore Technology Conference, Paper OTC 4321, Houston, TX. Boef, W. I. C., and Out, J. M. M., 1990, Analysis of a Flexible Riser Top Connection with Bend Restrictor, Proceedings Offshore Technology Conference, Paper OTC 6436, Houston, TX. Brooks, D. A., Kenison, R. C., and BP International Limited, 1989, Research & Development in Riser Systems, Subsea, London. Brown, P. A., Soltanahmadi, A. and Chandwani, R., 1989, Problems Encountered in Detailed Design of Flexible Riser Systems, Int. Seminar on Flexible Risers, University College, London. Claydon, P., Cook, G., Brown, P. A., Chandwani, R., Zentech International Ltd. London, 1991, A Theoretical Approach to Prediction of Service Life of Unbonded Flexible Pipes under Dynamic Loading Conditions, J. Marine Structures, preprint. Hoffman, D., Ismail, N., Nielsen, R., and Chandwani, R., 1991, Design of Flexible Marine Riser in Deep and Shallow Water, Proceedings Offshore Technology Conference, Paper OTC 6724, Houston, TX. Ismail, N. M.,1984, Wave-Current Models for Design of Marine Structures, Journal of the Waterway, Port, Coastal and Ocean Division, ASCE, Vol 110, No.4. Kastelein, H. J., Out, J. M. M., and Birch, A.D., 1987, Shell s Research Efforts in the Field of High-Pressure Flexible Pipe, Deepwater Offshore Technology Conference, Monaco. Kodaissi, E., Lemerchand, E., and Narzul, P.,1990, State of the Art on Dynamic Programs for Flexible Riser Systems, ASME Ninth International Conference on Offshore Mechanics and Artic Engineering, Houston, TX. Lopes, A. P., de silva Neto, S. F., Estefgn, S. F. and Da Silveria, M. P. R.,1990, Dynamic Behavior of a Flexible Line Design Installation, Proceedings Offshore Technology Conference, Paper No. OTC 6437, Houston, TX. 5

Machado, Z. L., and Dumay, J. M.,1980, Dynamic Production Riser on Enchova Field Offshore Brazil, Offshore Brazil Conference, Latin America Oil Show, Rio de Janeiro. Mahoney, T. R., and Bouvard, M. J., 1986, Flexible Production Riser System for Floating Product Application in the North Sea, Proceedings Offshore Technology Conference, Paper OTC 8163, Houston, TX. Oliveira, J. G., de Goto, Y., and Okamoto, T., 1985, Theoretical and Methodological Approaches to Flexible Pipe Design and Application, Proceedings Offshore Technology Conference, Paper OTC 5021, Houston, TX. Out, J. M. M., 1989, On the Prediction of the Endurance Strength of Flexible Pipe, Proceedings Offshore Technology Conference, Paper OTC 6165, Houston, TX. Sparks, C. P.,1983, The Influence of Tension, Pressure and Weight on Pipe and Riser Deformation and Stresses, 2nd International Offshore Mech and Artic Engineering Symposium,Houston, TX. Tillinghast, W. S.,1990, The Deepwater Pipeline System on the Jolliet Project, Proceedings Offshore Technology Conference, Paper OTC 6403, Houston, TX. Tillinghast, W. S. and Shah, B. C., 1987 Laying Flexible Pipelines Over Coral Reefs in the Geisum Field, Gulf of Suez, Egypt, Proceedings Offshore Technology Conference, Paper OTC 5585, Houston, TX. 6

Design data Deep water Shallow water steep wave Case A Free-Hanging Catenary Case B Easy-Touch Catenary configuration Flexible Riser Pipe Data: Internal diameter, M Outside diameter, M Axial Stiffness, N Bending Stiffness, Nm2 Wt in Air, kg/m Environmental data: Water depth, m Wave height, m Wave periods, s Current speed, m/sec Top, Bottom Vessel/Floater Excitation: Surge, m Heave, m 0.1016 0.1631 5.45E7 2.00E3 44.0 625.0 11.8 11.2 1.1,0.6 1.8 2.0 0.13 0.2 1.28E8 4.E3 62.0 350.0 17.0 12.0 1.9,0.7 2.2 5.5, -9.5 0.15 0.23 2.00E7 5.24E3 63.0 40.0 15.0 13.0 2.3,1.07 2.5 2.2 7

8

9

10

11

12

13

14