References, tables and figures at end of OTC 7297

Transcription

1 OTC 7297 Recent Developments in the Towing of Very Long Pipeline Bundles Using the CDTM Method Albertus Dercksen, Maritime Research Inst. Netherlands, and Taco Taconis, Rockwater Copyright Offshore Technology Conference This paper was presented at the 25th Annual OTC in Houston, Texas, U.S.A., 3-6 May This paper was selected for presentation by the OTC Program Committee following review of information contained in an abstract submitted by the Contents of the paper. as presented, have not been reviewed by the Offshore Technology Conference and are subject to correction by the The material, as presented, does not necessarily reflect any position of the Offshore Technology Conference or its offlcers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment of where and by whom the paper is presented. ABSTRACT Fully integrated pipeline bundles are built and tested onshore. The bundles are towed to the field using the Controlled Depth Towing Method (CDTM). Recently this method was used to tow a very long bundle (6.7 km) to the Piper field in the British sector of the North Sea. Time domain simulations were carried out to predict the bundle behaviour in different situations. A discussion of the numerical model will be presented. Several parts of the bundle configuration were model tested and full scale monitoring was performed during the tow at sea. Bundle tow in still water and survival sea states is discussed as well as the impacts of failure modes on the bundle behaviour. INTRODUCTION Fully integrated pipeline bundles are built and tested at the production site. After completion the bundle is launched and towed to the field. Often it will be necessary to cross other pipelines located on the tow route from the production site. References, tables and figures at end of paper. 355 To achieve clearance with these other pipelines it is essential to perform a safe tow. To this extent a dynamic analysis of the tow is carried out. The importance of having reliable numerical methods to predict the dynamic behaviour of towed pipeline bundles is given by the fact that it is extremely difficult to carry out model tests on these systems due to their immense lengths. The numerical model presented in this paper is based on a lumped mass approach, allowing flexibility with respect to modelling of different configurations. Model tests were carried out to gather information about essential items of the tow. Full scale monitoring was carried out to validate the numerical model for future projects. In principle CDTM involves the transportation of a pipeline bundle suspended between two tugs, the Leading Tug (LT) and the Trailing Tug (TT), Figure 1. To maintain control during tow, the bundle is designed and constructed within specific tolerances with respect to its submerged weight.

2 Recent Developments in the Towing of Very Long Pipeline Bundles Using the CDTM Method OTC 7297 The bundle is designed to have buoyancy, which is ensured by making the displacement of the bundle greater than its dry weight. Ballast chains are attached to the bundle in order to give it a small submerged weight and also to provide lift forces during tow. Both ends of the bundle are equipped with a towhead structure. When commencing tow, the bundle is tensioned by the tugs and the pull force of the leading tug is gradually increased and set to a value pertaining to the required tow speed and the pull force of the trailing tug. The increasing tow speed results in increasing lift forces on the bundle and finally the ballast chains are lifted free from the seabed. The tow speed has a direct lift and straightening effect on the bundle, so by controlling the tow force of the leading tug in combination with the pre-tension maintained by the trailing tug, bundle deflections are kept under control. By continuously monitoring the bundle during tow, the deflections are kept within operational limits. Upon arrival in the field, the bundle is gradually lowered by pre-determined adjustment of the tow wire length and tensions. The bundle settles in a position above the seabed with the ballast chains resting partly on the seabed. From this position, the bundle can be put into its correct location with the towheads in the required target areas. Specific attention is paid to the maximum bending moments, deflections and line tensions during various tow situations. The main purpose of the time domain analysis is to- determine the behaviour of - the bundle during tow under severe conditions. Further the bundle behaviour as a result of a tow line failure is investigated. THEORY method". This technique involves the lumping of mass, excitation forces due to wave and current, and reaction forces (fluid reactive forces, tension, shear forces and soil forces) at a finite number of nodes along the pipeline. All element forces are formulated in terms of element properties, position and orientation. By applying the equations of dynamic equilibrium and stress-strain compatibility to each node, a discrete set of equazions of motion is derived. These equations are solved in the time domain by use of finite difference techniques and iteration procedures. The space-wise discretization of the line consists of concentrated masses interconnected by massless bars which act as rigid beams in normal direction and as linear springs in axial direction, see Figure 2. The element tension is a linear function of the instantaneous distance between the adjacent nodes. Bending moments are derived from the element orientation~. It is assumed that the elements have axi-symmetrical (cylindrical) properties as to the dimensions, fluid force coefficients and stiffness. The governing equations of motion for the j-th lumped mass, Newton's law, is written in global coordinates, ([Aj] + [a.]) Y.(t) = F.(t) J J J where : [A.] = inertia matrix [a?] = time dependent added inertia matrix t = time X =.. j position vector X. = acceleration vector F' = nodal force vector. j - The added inertia matrix can be derived from the normal and tangential fluid inertia coefficients by directional transformations. The nodal force vector contains the following contributions: Equations of motion - segment tension T - buoyancy and weight - The simulations are carried out using the fluid fbrces multi-functional time domain simulation seabed reactive forces program DYNFLXf90, ref. [l]. - The matheshear forces due to bending rigidity matical model is known as the "lumped mass - user supplied forces.

3 OTC 7297 A. Dercksen and T. Taconis 3 Hawser modelling Direct coupling of tug motions and bundle motions are suppressed by incorporating nylon hawsers, acting as stretchers, in the tow lines. The non-linear stretch properties of the hawsers are modelled separately. From the relationship between the tension T, elongation E and the breaking strength T, Figure 3, a formulation for the stif'fness EA as a function of the tension and ultimate break load is determined: in which a, p and y are coefficients depending on the type of hawser. At each simulation time step the stiffness of the element corresponding to the hawser is updated according to the above formula. Chain lift and drag forces Typical for CDTM applications is the presence of ballast chains to generate lift forces which can be controlled by the towing tugs. Experimental data on the lift and drag forces of chains was obtained from full scale model tests. The following formulas were used: in which L is the lift force, D the drag force, V the tow speed, see Figure 4, and yl, ay p1 and p2 are coefficients depending on chain length, chain type and velocity range. The parameters were obtained from the model test results by means of regression. The prototype bundle has ballast chains at fixed intervals along the bundle. In the simulation, the resulting lift and drag contributions are incorporated in the nodal forces. Seabed reaction forces The part of the ballast chains which is lying on the seabed will undergo seabed friction, resulting in a friction force opposite in direction to the velocity of the chain, and proportional to the weight of the part lying on the seabed. The friction coefficient for the ballast chains was 0.6, for the tow wire it was 1.0. As the bundle moves closer to the seabed, the ballast chains start acting as springs. At a given tow force, an equilibrium situation occurs at which the bundle remains at a certain height above the seabed, Figure 5. This situation occurs at the off-bottom tow at low velocities, and after a leading tow line failure. Towhead forces The pipeline bundles are equipped with structures at each end, in this context referred to as towheads, Figure l. The towheads are modelled as tubular elements with modified fluid coefficients in order to represent the correct rowhead loads. Two different towhead models were towed at several velocities and angles of attack, The measured forces were made dimensionless to obtain the hydrodynamic coefficients as functions of towhead yaw and pitch angles. The actual towhead forces were calculated by means of the relative motion concept. motions Tug motions in waves were obtained from the tug response amplitude operators and wave records. In this way both regular and irregular motions can be inputted to the simulation. Manoeuvring of the tugs can be incorporated by superimposing a certain track, or by specifying an external force acting on the tug. EXPERIMENTS As mentioned in previous sections, model tests were conducted at the Maritime Research Institute Netherlands. The main purpose of the tests was to obtain input data for the simulations. Two separate programs were carried out. Towhead model tests A towhead model was built to scale 20. Both ends of the bundle were equipped with a practically identical towhead structure.

4 Recent Developments in the Towing of Very Long Pipeline Bundles Using the CDTM Method OTC 7297 Figure 6 shows the instrumented model of the towhead. The model was mounted an a 5-component force transducer. Tests were carried out in still water with the towhead fully submerged; tow speed, direction, pitch and yaw angles were varied. The tests showed the following: - longitudinal forces are approximately independent of the pitch or yaw angle variation; - vertical forces and pitch moments are strongly dependent on the pitch angle; - transverse forces and yaw moments are strongly dependent on the yaw angle. Chain model tests The chain model tests were carried out using full scale chain segments. The segments were connected to a 6-component measurement frame which in turn was connected to the towing carriage. The tests involved 2 types of chain (76 mm and 64 mm), each chain was tested at 3 different lengths (8, 10 and 12 links) and the heaviest chain was also tested with a wrapping material. The results were analysed and the following conclusions were drawn: - Fitting wrapping material around vertically hanging submerged chains appears to increase the lift coefficient with a factor of 3 and the drag coefficient with a factor of 2. The lift/- drag ration increases roughly with a factor 1.5, Figure 7. - The length of the chains has only a minor effect on the dimensionless lift and drag coefficient. - Different wrapping or damaged wrapping did not cause significant variations in the lift and drag coefficients of the chains. FULL SCALE DATA Full scale measurements during the bundle tow are limited. Main attention focuses on the bundle depth, especially when passing other pipelines. In high sea states the towheads and bundle are kept well underwater to minimize wave action on the bundle. From the data logs of the actual tow, it was concluded that the average tow speed was between 10 and 20 percent higher than the tow speed used in the simulation~. The total drag was comparable with the simulations. Explanations for the difference are a conservative longitudinal drag coefficient on the bundle. Weight variations along the bundle require locally higher or lower lift from the chains and external factors such as varying current velocities along the bundle, boundary layers along the bundle and sea water density variations. SIMULATIONS Simulations were carried out for several bundles. In this paper the longest bundle ever towed is discussed. Bundle particulars are presented in Table 1. The bundle configuration including tow lines was represented by 95 nodes, 94 elements according to Table 2. The bundle was towed by two leading tugs, in the simulations they were modelled by one equivalent tug. The static tow configuration is defined as the shape of the bundle including chains, towheads and tow lines at a constant speed in calm water. The forward speed was modelled by fixing the tugs in space and applying a constant current of approximately 1.6 m/s. The resulting static tow configuration is presented in Table 3, and Figure 8. Prior to parking the bundle in its required target location, the bundle must be lowered to the seabed. This is accomplished by paying out the tow line and reducing the tow speed. The resulting offbottom tow is shown in Figure 9. An important failure mode of a tow is the break of the leading tow line. The situation was modelled by setting the axial stiffness of the element attached to the leading tug to a very small value. The trailing tug was required to maintain a

5 OTC 7297 A. Dercksen and T. Taconis 5 constant pull force. As a result of this the trailing tug accelerates in a direction opposite to the tow direction. The leading tow line will fall to the seabed pulling the leading end of the bundle down. Due to the reduced speed, the entire bundle slides into an equilibrium configuration. A geometric presentation of the bundle at different time intervals is shown in Figure 9. Another failure situation is the break of the trailing tow line. During the simulation, the axial stiffness of the element attached to the trailing tug was set to a very small value, and the leading tug maintained a constant pull force. Due to the decreased resistance, the leading tug then accelerates and the bundle eventually appears at the surface due to the increased lift forces. By reducing the tow force of the leading tug, the bundle surfacing can be prevented. The surfacing of the bundle is graphically presented in Figure 11. Tow during survival conditions was simulated by towing at a constant velocity in a Beaufort 8 sea state. Tug motions were prepared off-line and the influence of bundle behaviour on the tug motions was neglected. Since the towheads and bundle are well below the still water level, and the tow lines contain nylon stretchers, the bundle deflections are not significantly larger than the static deflection, Figure 12. Peak loads in tow line tension and bundle bending moments increase considerably as is shown in Table 4. DISCUSSION The simulations as described in the previous section were carried out in idealized situations. For design purposes, it is required that the simulations do not underestimate the bundle behaviour. The static tow configuration is merely a well-defined starting point for the simulation~. In reality the Tow Master will continuously monitor the bundle depth and position and take corrective actions, e.g. course changes to keep on track, speed loss due to waves or wind. The situation in which the tow speed and the distance between the tugs are kept constant as is modelled in the simulation will not occur often. The failure modes in which either the leading or trailing tow line break will most likely occur is during survival conditions as these generate the highest loads in the system. The bundle behaviour after a tow line breakage will not be essentially influenced by the environmental conditions. If the leading tow line breaks, the bundle will end up on the seabed if no corrective actions are taken. At a break of the trailing tow line, the bundle will surface if the pull force of the leading tug is not decreased. The dynamic bundle behaviour in waves was simulated under the assumption that the tug motions were not influenced by the bundle. In practice the tug motions may be influenced by the bundle and there may be an auto-tensioning device on the leading tug controlling the tow line tensions. The dynamic tensions resulting from the simulation~ are therefore conservative. Comparison of the simulation results and full scale data is required to improve on the modelling in the simulations and obtain information on the reliability of the results. In order to make a good comparison, it is essential to have a good representation of the tow at various instances, i.e. the following information should be measured if possible: tow speed - tow line tensions/tow force - bundle depth at several locations on the bundle tug motions bundle accelerations sea water density - relative velocity at several locations along the bundle - wave height/direction current speed/direction - wind speed/direction. The first three items are at present monitored for each tow.

6 Recent Developments in the Towing of Very Long Pipeline Bundles Using the CDTM Method OTC 7297 CONCLUSIONS Conclusions to be drawn from the simulations on the towed pipeline bundle can be summarized as follows: - Bundle behaviour is rather dependent on local velocities on the ballast chains. Wave frequency tug motions have relatively small influence on the deflections and bending moments. - Maximum bending moments are exerted at the towheads, especially at the leading towhead. - To prevent surfacing of the bundle to occur, corrective actions are required after a break of the trailing tow line. - For better understanding and improvement of the simulation models, full scale measurements are required. - Future projects may involve even longer pipeline bundles. If the accuracy of the simulations is known, the feasibility of the project may be determined with the aid of simulations. To this extent validation material is required. REFERENCES 1. Boom, H.J.J. van den, Dekker, J.N. and Elsacker, A.W. van; "Dynamic Aspects of Offshore Riser and Mooring Concepts", OTC Paper 5531, 1987.

7 OTC 7297 A. Dercksen and T. Taconis 7 Table 1. Particulars of bundle Mass Length Outer diameter Axial stiffness Bending stiffness E10 l. le W Layout : 1 gas export line 16" 1 oil export line 10" 1 gas lift line 8" 1 water injection line 16" ---h Towheads : LTH and TTH of 22 m with integrated Underwater Safety Valves [kg/ml [m] [m] [NI2 [Nm ] Table 3. Results of static tow analysis Tow velocity Tension TT Tension LT Depth of bundle Depth TTH Depth LTH [mls] [W1 fw tml fml [m] Table 2. Discretization of bundle configuration Description Tow wire 56 mm Hawser Bridle + pennant Trailing towhead Bundle Leading towhead Bridle + pennant Tow wire 80 mm Hawser Tow wire 76 mm Length [m] Element lengths [m] 2x x x23,2~30,6~40,6~50, 4x75,57~ ,2x40 3x x50,48 Table 4. Comparison of static tow and tow in survival conditions Description Static tow minimum Dynamic tow maximum Unit Depth of bundle Depth TTH Depth LTH Tension TT Tension LT Bending moment [m] [m] [m] [knl /m1 [mm]

8 Recent Developments in the Towing of Very Long Pipeline Bundles Using the CDTM Method OTC 7297 Trailing tug (TT] Leading tug (LT) Trailing towhead Leading towhead g (TTH Ballast chains (LTH) T \\V/fA\V//.5\\7/// E l Fig. 1 CDTM configuration Fig. 2 Lumped mass method

9 OTC 7297 A. Dercksen and T. Taconis 9 Fig. 3 Lqad characteristics of nylon hawsers

10 Recent Developments in the Towing of Very Long Pipeline Bundles Using the CDTM Method OTC 7297 Lift Tow speed 7 Weight Fig. 4 Drag and lift forces on chains during tow Seabed \\V//A\Y//AV//A\\\ Fig. 5 Chain configuration near the seabed

11 OTC 7297 A. Dercksen and T. Taconis Fig. 6 Scale 1 to 20 model of TOWHEAD 1.o Li ft/drag..-.. Bare chain Wrapped chain Tow speed in m/s Fig. 7 LiftlDrag ratio for bare chains versus wrapped chains

12 Becent Developments in the Towing of Very Long Pipeline Bundles Using the CDTM Hethod OTC Horizontal distance along bundle [m1 Fig. 8 Static tow configuration 150 Waterdepth 148 m f.- O I I I Horizontal distance along bundle [m1 Fig. 9 Off bottom tow configuration 366

13 OTC 7297 A. Dercksen and T. Taconis Fig. 10 Break of leadind yow line Fig. 1 1 Break of trailing tow line

14 Recent Developments in the Towing of Very Long Pipeline Bundles Using the CDTM Method OTC 7297 TRAILING TUG I I I TRAILING TOWHEAD LEADING TOWHEAD l 1 " " 1 " " 1 ' " ' 1 " " I " " I " " I " " I ' " ' I " ~ ' l SECONDS Fig. 12 Tow in survival conditions, vertical motions, wave direction 180 degrees