Proceedings of the ASME nd International Conference on Ocean, Offshore and Arctic Engineering OMAE2013 June 9-14, 2013, Nantes, France

Transcription

1 Proceedings of the ASME nd International Conference on Ocean, Offshore and Arctic Engineering OMAE213 June 9-14, 213, Nantes, France OMAE SIMULATION OF HOOKING EVENT IN FISH TRAWLING OPERATION Xiaopeng Wu Centre for Ships and Ocean Structures(CeSOS) Norwegian University of Science and Technology NO-7491, Trondheim, Norway Vegard Longva Department of Marine Technology Norwegian University of Science and Technology NO-7491, Trondheim, Norway Svein Sævik Department of Marine Technology Norwegian University of Science and Technology NO-7491, Trondheim, Norway Torgeir Moan Centre for Ships and Ocean Structures(CeSOS) Norwegian University of Science and Technology NO-7491, Trondheim, Norway ABSTRACT Interference between trawl gears and subsea pipelines is an important issue. A special case called hooking, defined as the situation when the trawl gear gets stuck under the pipeline, should be a rarely occurring situation. In this case, however, the warp line tension could be high as its breaking strength. This may be detrimental with respect to both fishing vessel safety and pipeline integrity. This calls for a better understanding of the hooking phenomenon. The goal of this study is to develop a proper numerical model to describe the hooking event. The proposed model is based on the finite element method. A special penalty-based contact element that includes the friction effect is utilized to deal with the trawl board and pipeline interaction. The trawl board and seabed (span shoulder) contact is also accounted for in order to simulate the hooking event. To validate the proposed model, numerical simulation results are compared with previous model test results. A rectangular type trawl board was selected as the target object. The pull-over cases with different span height were firstly tested and compared. Then, the hooking event set-ups were modeled. Based on the model tests, there are two most likely scenarios for hooking: 1) a de-stabilized trawl board with small span height; 2) small crossing angle with large span height. The above two cases were both tested by the proposed model. In the first case, the trawl board is towed flat on the seabed. Permanent hooking was successfully obtained in the simulation. Then, two cases with small crossing angle were studied. Hooking event was reproduced in the case of a 2 degree crossing angle by introducing a disturbance on the trawl board. It shows that the proposed model could reproduce the hooking event, provided that the trawl board motion similar to the model test could be obtained. This gives a good basis for further studies. INTRODUCTION It is a well-known fact that subsea facilities attract fishes. So the conflict between offshore oil industry and seafood production activities could not be eliminated. One potential risk is that bottom fish trawling gear may collide with subsea pipelines. According to the DNV-RP-F111 [1], this kind of interference could be classified into two phases, the initial impact and pull-over. During the impact phase, trawl gear exerts a short but powerful impulse force on the pipeline, which could cause local coating damage or even denting on the pipe wall. The pull-over event follows the initial impact subsequently, which usually gives a more 1 Copyright 213 by ASME

2 global response of the pipeline. Many researchers have put their effort into investigating the initial impact as well as the pull-over phenomenon. Kjeldsen and Moshagen [2, 3] investigated the influence of the bottom trawl gear on the pipeline with fixed end condition through model test as early as in the 197s. More experiments were carried out by MARINTEK in 199 [4], in which the pipeline end flexibility was taken into account. Verley [5,6] focus on the pull-over force time-history for trawl boards, while Guijt and Horenberg [7] studied on beam trawls pull-over. There is also a rarely happened but hazardous case in trawl gear-pipeline interaction, which is the so-called hooking event. It is defined as the trawling equipment gets stuck under the pipeline. In this situation, the warp line tension could reach as high as its breaking strength (typically in the order of 4 tonnes), and this force will be acting on the pipeline. Buckling of crosssection could happen due to excessive bending, especially for small diameter pipe. Hooking could also cost lives. In 1997, a fishing vessel, Westhaven, capsized with four casualties as the crew attempted to free the hooked port trawl board from a 3- inches pipeline [8]. The possible severe consequence caused by the hooking event calls for better understanding of the phenomenon. Hooking event is most likely to occur in two situations, according to the observation in the experiment and the DNV Rules [1]. 1) A de-stabilised trawl board (dragged on its back along the seabed) approaches a pipeline with a small gap. The trawl board may dig under the pipeline and get hooked. 2) The crossing angle with pipeline is less than 45 deg and the pipeline is free spanning. The warp line lifts the trawl board off the seabed, it slides along the pipeline, becomes de-stabilised, turns over and slides underneath the pipeline until it gets wedged at the span shoulder. A typical hooking scenario at the free span is illustrated in Fig. 1. In the DNV document DNV-OS-F11 [9], hooking is categorized as an accidental load, since it has a small occurring frequency per year. The hooking response of pipeline is treated as a static lifting up procedure in the recommended practice [1]. A specific lifting height, might be limited by the maximum warp line force, is applied on the pipeline in combination with other possible restraint. Then all relevant failure modes shall be checked. This method deals only with the consequence of the hooking. However, if one wants to reduce the probability of hooking, different approaches can be applied. Model testing has been proved to be studying such effects. Here, an alternative way is applied, developing a proper numerical model which could reproduce what happened in the laboratory. With a validated model, sensitive parameters in the hooking event can be investigated, and more insight can be gained. As a result, better design of trawl equipment, pipeline facilities or even fish trawl- FIGURE 1. SKETCH OF TYPICAL HOOKING SCENARIO IN A FREE SPAN [1] ing procedures may be proposed. OBJECTIVE The objectives of this paper are as follows: 1. To propose a numerical model with good efficiency and potential to study the hooking event. 2. To reconstruct the hooking test set-ups and reproduce the hooking scenario with the proposed model and to compare the simulation results with model test results. NUMERICAL MODEL DESCRIPTION The SIMLA [1] code is a finite element code which is specially designed to handle static and dynamic problems of long slender structures, i.e. pipelines. Different non-linearities such as geometric, material and contact could be handled in this code. The numerical simulation in this study is performed by SIMLA. There are two important aspects relating to the hooking event, the hydrodynamics loads on the trawl gear, and the contact problem involved (between trawl board and pipeline or/and seabed). The hydrodynamic forces acting on the trawl gear are related to its orientation as well as the relative velocity and acceleration of the ambient fluid. They are also influenced by the seabed or pipeline proximity. A proper description of the hydrodynamics of the trawl is of significance to simulate its motion in the water. Besides, the contact problem is a key issue in a hooking event. In different scenarios, the trawl could get hooked 2 Copyright 213 by ASME

3 due to the contact with different objects, such as seabed (span shoulder), pipeline and possible mechanical deflector. The contact mechanism like friction and penetration should be described in a decent manner in order to reproduce the hooking event. The numerical modeling considerations in this study will be briefly addressed in the following sub-section. Hydrodynamic Model The trawl board is represented by a rigid body motion model of 6-DOF. The hydrodynamic loads acting on the trawl board are according to Eqns. (1) - (3). F A = m a11 F D2 = 1 2 ρ m a22 SY M. m a33 m a44 m a55 m a66 η 1 η 2 η 3 η 4 η 5 η 6 (1) F D1 = 1 C 2 2 ρ C 4 V 2 HL (2) C 6 C 11 C 22 SY M. C 33 C 44 C 55 C 66 η 1 η 1 η 2 η 2 η 3 η 3 η 4 η 4 η 5 η 5 η 6 η 6 Equation (1) expresses the hydrodynamic mass loads, in which m aii denotes the added mass coefficient in the i th DOF, η i denotes the acceleration of the trawl board in the i th DOF in the local coordinate system. The hydrodynamic mass coefficients depend on two variables, the gap between the trawl board and the seabed, and the trawl board inclination angle with respect to the seabed. These coefficients were calculated by Statoil for a rectangular plate geometry with an aspect ratio of.76 resting on the seabed. Panel method was used in their calculation. Further detail is found in [11]. Equation (2) describes the drag load on the trawl due to forward speed, in which ρ is the water density, C i is the drag coefficient in the i th DOF, V is the global horizontal velocity vector (3) Drag Force Lever Arm l [ ] L/4 1 L/8 L/8 L/ Trawl Board Heading Angle θ [ ] FIGURE 2. DRAG FORCE LEVER ARM WITH RESPECT TO TRAWL BOARD HEADING ANGLE of the trawl board, H and L are the height and length of the trawl board, respectively. C 1, C 3 and C 5 are considered to be small in most cases and are thus set to zero. C 2 is estimated based on a thin flat plate inclined to flow in [12] and see Eqn. (4), { Γ2π tanθ,θ < 8 C 2 (θ) = Γ /sinθ,9 12 (4) in which θ is the heading angle of the trawl board, Γ is a knock-down factor to account for 3D effects and set to.465 in this study. C 4 is properly set so that the trawl board will not be uplifted during the acceleration phase. C 6 is calculated as Eqn. (5). C 6 = C 2 (θ)l(θ) (5) in which l is the lever arm of the drag force, i.e. the distance between the hydrodynamic drag center and the geometrical center of the board. According to the lifting surface theory, the drag force center will be located a quarter of the cord length from the front edge when the heading angle is small. Since the drag center of a flat plate when it is perpendicular to the flow will be at the geometric center, a simple relation between the lever arm of the drag force (in terms of trawl board length L) and the trawl board heading angle is then assumed, see Fig. 2. Equation (3) calculates the drag loads on the trawl due to local velocities. They are important during the pull-over phase because the forward speed in this phase is relative small, whereas local rotation angular velocities could be large. 3 Copyright 213 by ASME

4 FIGURE 3. 2r BODY GEOMETRY ELEMENT Contact Element The contact problem in this study is handled by a bodypipeline contact element based on the penalty method. Different from the conventional node-to-surface contact formulation, an edge-to-beam formulation is implemented. In this way, the contact search becomes more efficient. The body is assumed rigid such that no information about the trawl gear deformation can be defined. This is considered acceptable within the scope of this paper. The trawl board is considered to be a rigid body. Triangle plate elements are used to represent its geometrical shape. The triangle element consists of a plane triangular plate with thickness 2r surrounded by circular-shaped edges and sphericalshaped corners with radius r. And the r is the same among all the elements that build up the object, so that a smooth surface of the body could be obtained, see Fig. 3. The flat plate part is excluded from the contact geometry, since the pipe elements are assumed to be approximately straight. Thus, the contact geometry consists of circular-shaped edges and spherical corner surfaces, which are in green in Fig. 3. Then the contact search could be performed by measuring the minimum distance between a pipe element and the outer surface of the body. If the gap turns out to be of negative value, contact is defined to occur. Then the relevant contact force and subsequent friction force will be calculated. For more details, see Longva [13]. HOOKING EVENT SET-UP As what has been mentioned in the introduction, two typical scenarios may lead to a hooking event. They are when the trawl board is de-stabilized with small span height, and when the trawling direction and the pipeline orientation is less than 45 degree. In order to validate the ability of simulating hooking event with the proposed numerical model, the above two scenarios should be tested. There was an experiment funded by Statoil and conducted by MARINTEK in 22, see [14], with the aim of gas pipeline overtrawlability. One goal of this test was to find out the potential of hooking, so several hooking scenarios were constructed, including the two above scenarios. The simulation results can then be compared with the model test and the performance of the proposed model can be validated. TABLE 1. MODEL PROPERTIES Quantity Symbol Value Unit Water depth d 31.2 m Water density ρ 1 kg/m 3 Warpline stiffness 28 kn/m Warpline tension 4(flat)/6(other) kn Warpline angle θ W 18.3 deg Sweep line angle θ S 3 deg Trawl net resistance 4 kn Trawling velocity V 2 m/s Pipeline diameter.75 m Pipeline span height H SP.5,1,5 m Span shoulder slope β 3 deg Crossing angle α 9,45,2 deg The numerical model is based on Longva s work [13]. A rectangular type of otter trawl board was selected for the test. Firstly, the pull-over scenarios with different span height comparing the behavior of the trawl during the pull-over process was performed. This was done by a comparison between the simulation results and the recorded test video. Then, the three special cases in which hooking event occurred were studied. The numerical modeling set up is illustrated in Fig. 4. It consists of a trawl board of port side, one warp line with the top end connecting to towing node, two sweep lines and one node representing the trawl net. The starboard trawl board is represented by a node moving at the same speed as the towing node. A rigid pipeline section is modeled with different span height (H SP ). The seabed is modeled as a flat surface. And a flat span shoulder with 3 degree slope (β), perpendicular to the pipeline direction, is modeled according to the model test. The water depth (d) is set to 31.2m. Relevant quantities used in the numerical model are shown in Tab. 1. The pipeline section in the model is fixed in both ends. It is rigid with a diameter of.75 m and a length of 3m. The span shoulder on the starboard side is placed underneath the pipeline to represent a slope configuration. Since the contact element used in this study could only handle the contact between object and cylindrical sections, the span shoulder configuration is represented by a pipe section with very large diameter (1 m) and proper material properties. The span shoulder was rotated together with the pipeline section when the crossing angle (α) was adjusted. The warp line stiffness is 28 kn/m. It is selected to give reasonable warp line stiffness when the water depth is 35 m and the warp line length is approximately 3 times the water depth in full scale. In the model, only the lower part of the warp line is modeled and with a warp line angle (θ W ) of 18.3 degree in the vertical plane. Due to the possibility of interaction with pipeline, the warp line is modeled by an elastic stiffness corresponding to 4 Copyright 213 by ASME

5 Z X Y Trawl net node θs Sweep line Trawl board Starboard trawl board node θw Warp line Towing node d Trawling path Fixed Hsp Pipeline α Fixed β Span shoulder Slope FIGURE 4. MODEL CONFIGURATION a 28 mm steel wire and element lengths of 1 mm in the lower part. The model includes also the sweep lines and one node representing the trawl net. Drag coefficient is assigned to the trawl net node. The sweep line angle (θ S ) as well as the drag coefficient has an influence on the trawl board back tension. By adjusting these, proper trawl board motion could be obtained during the acceleration phase as well as the stabilization phase. The warp line tension level was adjusted in the same way. For example, in the de-stabilized case, the back tension should be lower, and thus the drag coefficient was reduced. The towing node moves with a prescribed displacement which gives a velocity of 2 m/s after about 9 seconds. The starboard trawl board node moves forward at the same speed as the towing node. It represents a trawl board motion without interference, such that a correct back tension level (close to zero) for the port side trawl board is obtained when pull-over or hooking occurs. The trawl board used in the model is a rectangular type trawl board, which has a steel mass of 26 kg in full scale with a height of 2.2 m and a length of 4. m. This is the same shape as used in the model test with a slightly difference in mass (25 kg in full scale). The trawl board is modeled by 236 triangular plates. The contact element mentioned in previous section was applied. The stiffness of the contact element was tuned to represent realistic contact behavior between the trawl board and the rigid pipeline and the span shoulder. The mass moments of inertia of the trawl board are given in Tab. 2. TABLE 2. TRAWL BOARD PROPERTIES Quantity Value Unit Height 2.2 m Length 4 m Mass 26 kg C. O. G. x-coordinate -.2 m C. O. G. z-coordinate -.11 m Roll mass moment of inertia 1885 kgm 2 Pitch mass moment of inertia 5245 kgm 2 Yaw mass moment of inertia 3358 kgm 2 response in the videos. Pull-over Cases With Different Span Height Several simulations of pull-over cases with different span height were compared against the model test results. In these cases, the trawl board had a velocity of 2 m/s with the trawling path perpendicular to the pipeline section (α = 9 ). No span shoulder was present in these cases as there is no contact between the trawl board and the span shoulder. Cases of three span heights (H SP ) were tested:.5m, 1m and 5m. The warp line time histories for these cases are shown in Fig Only 2 seconds of time history within which the pull-over occurred are presented. The maximum warp line tension and pull-over duration are tabulated in Tab. 3. The snap shot series of both simulation visualization and model test are presented in the Appendix A. RESULTS AND DISCUSSION Since the main purpose of the model test is to visualize the overtrawlability processes, just the warp line tension and videos were recorded in the test. No detail measures were taken on the pipeline impact force or trawl board motions. It is therefore important to compare the simulation results with the test with respect to the warp line tension time history and the trawl board TABLE 3. SUMMARY FOR PULL-OVER CASES Case Maximum Tension [kn] Pull-over Duration [s] Simulation Model Test Simulation Model Test H sp =.5m H sp = 1.m H sp = 5.m Copyright 213 by ASME

6 Warp Line Tension [kn] Simulation Model Test Time [s] FIGURE 8. PULL-OVER SNAP SHOT, SIMULATION, H SP = 5.m, α = 9 FIGURE 5. WARP LINE TENSION, H SP =.5m, α = 9 Warp Line Tension [kn] Simulation Model Test FIGURE 6. Warp Line Tension [kn] Time [s] FIGURE 7. WARP LINE TENSION, H SP = 1.m, α = 9 Simulation Model Test Time [s] WARP LINE TENSION, H SP = 5.m, α = 9 In general, when the span height is small, the simulation re- FIGURE 9. PULL-OVER SNAP SHOT, MODEL TEST, H SP = 5.m, α = 9 sults give reasonable agreement with the model test results. The maximum tension in the warp line is of the same level when span height is.5 m and 1. m. However, in the case of 5. m span height, the warp line tension in the simulation is underestimated with significant discrepancy. One simulation snap shot for the case of 5. m span height is shown in Fig. 8. By comparing it with the snap shot from the video in the test, Fig. 9, one difference is observed. The trawl board in the test rotates mainly around the pipeline section with its lower front part (Fig. 9), while the one in the simulation rotates mainly with its lower side part (Fig. 8). This behavior was initiated since the uplifting process began when the warp line had contact with the pipe. In the test the trawl board is observed to rotate around its longitudinal axis to a more horizontal position, while the trawl tends to keep its inclination angle in the simulation. Due to this difference, the duration of the pull-over in the test is one second longer than that in the simulation, and thus a larger warp line tension is measured. One possible reason for this may be that the hydrodynamic load induced by vertical motion is not included in this model. 6 Copyright 213 by ASME

7 Warp Line Tension [kn] FIGURE 1. Simulation Model Test Time [s] WARP LINE TENSION, H SP =.5m, α = 9, FLAT Hooking Events There are several cases that hooking event occurred in the model test. Three of them were selected to be reconstructed in the proposed numerical model, including one case with destabilized trawl board and two with crossing angle smaller than 45 degree. In the model test, the breaking strength of the warp line is set to 5 kn, which is reasonable as the value is usually 4 kn in the real case. If permanent hooking occurred in the test, a safety spring will be active when the warp line tension exceeds 5 kn. Information about the hooking scenarios simulated are tabulated in Tab. 4. TABLE 4. SUMMARY FOR HOOKING CASES Hooking H SP α Span Hooking in Hooking in Type [m] [ ] Shoulder Model Test Simulation De-stabilized.5 9 Without Yes Yes Small α With Yes No Small α 5. 2 With Yes Yes De-stabilized Trawl Board With Small Span Height In the de-stabilized case, the trawl board was dragged on its back along the seabed. This was achieved by reducing the drag coefficient in the trawl net node in the numerical model. The warp line tension time history is shown in Fig. 1. The snap shot series of both simulation visualization and model test are presented in the Appendix B. Permanent hooking was successfully reproduced in this case. The trawl board was hooked with its front edge pushing against the seabed while the connection part was pulled against the pipeline section. The warp line tension exceeded 4 kn before the simulation stopped. In the model test, the towing carriage slowed down a little bit earlier than other set-ups to protect the warp line model from breaking, this is why the maximum warp line tension measured was lower than 35 kn in the model test. This case shows that the pipeline-body contact element is working well. Small Crossing Angle With Large Span Height When the trawling crossing angle is less than 45 degree, the trawl board will tend to slide along the pipeline rather than be pulledover. For certain types of trawl board, if the sliding occurs near the span shoulder, the probability of hooking will be increased. That is well illustrated in the test, as the MALO type trawl board gets stuck in both 45 and 2 degree crossing angle cases. However, the final hooking position is different between these two cases, based on observation. When the crossing angle is 45 degree, the trawl board was hooked as its lower front edge gets stuck between the pipeline and the span shoulder, see Fig. 11(a). In the case with 2 degree crossing angle, however, the trawl board was stuck with its upper and lower side edges leaning against the pipeline and the span shoulder, respectively, see Fig. 11(b). This is because the way of the trawl board approaching the pipeline is different from each other. In the first case, the trawl board hits the pipeline first and then slides into the span shoulder corner. In the following case, the trawl approaches the pipe line with its lower front edge resting on the span shoulder. With the proposed model, effort has been put into simulating the two cases mentioned above. The span shoulder perpendicular to the pipeline section was reconstructed in both cases. It was represented by a large pipe section with a diameter of 1 m. The warp line tension time histories for the two cases are shown in Fig. 12 and Fig. 13. The snap shot series of both simulation visualization and model test for the 2 degree case are presented in the Appendix B. As illustrated in Fig. 12, with a crossing angle of 45 degree, the proposed numerical model failed to reproduce the hooking event as what occurred in the model test. In the simulation, the trawl board kept its inclination angle large during the up lifting phase, while the inclination went smaller in the model test. This kind of behavior is quite similar with the case of the same span height, but of a crossing angle of 9 degree. Then, the trawl board slides into the span shoulder corner without a correct position and was not hooked. In the case of 2 degree crossing angle, the motion of the trawl board is more similar to the one in the model test than the first case. However, the trawl does not get hooked in the simulation. By detailed observation of the model test video and analysis on the simulation results, it was found that the trawl board was not approaching the pipeline close enough. There was also a subtle discrepancy in the tilt angle observed. In order to reproduce 7 Copyright 213 by ASME

8 Warp Line Tension [kn] Simulation Model Test Time [s] (a) α = 45 FIGURE 12. WARP LINE TENSION, H SP = 5.m, α = 45, WITH SPAN SHOULDER (b) α = 2 FIGURE 11. FINAL HOOKING POSITION, MODEL TEST, H SP = 5.m, WITH SPAN SHOULDER approximately the same motion as that occurred in the model test, a small disturbance was introduced on the trawl board before it hitted the pipeline. A horizontal force pushes the trawl a little bit more underneath the pipeline, while a moment adjusts the pitch angle of the trawl. The force and moment were ramped up and down to avoid significant transient effect. The magnitudes were chosen as the lowest level that could reproduced the hooking. The hooking event was then reproduced in the simulation. The warp line tension time history is in good agreement as shown in Fig. 13. This case shows that the proposed model could reproduce the hooking scenario, provided that the trawl board motion similar to the model test could be obtained. The 6-DOF motion of the trawl in the few seconds before the hooking event occurs is crucial. Warp Line Tension [kn] Simulation Model Test Time [s] FIGURE 13. WARP LINE TENSION, H SP = 5.m, α = 2, WITH SPAN SHOULDER CONCLUDING REMARKS AND FUTURE WORK In this paper, a numerical model with a penalty-based contact element is proposed to study the hooking event in fish trawling operation. Simulation results were compared with existing model test results. Three hooking scenarios, as well as three pullover cases in the model test were reconstructed. It is encouraging that the case with a de-stabilized trawl board was successfully reproduced, which shows that the contact element is working well. In one case with small crossing angle (α = 2 ), hooking is captured by introducing a disturbance before the hooking occurred. It demonstrates that the hooking event could be reproduced by the model, provided that a realistic trawl board response is obtained. In future work, a more detailed hydrodynamic model is needed to describe the subtle loads of trawl, especially the forces during the up-lifting period. The relative vertical velocity in- 8 Copyright 213 by ASME

9 duced by the trawl board response might play an important role here. The effect of span shoulder on the hydrodynamic loads on the trawl should be investigated, too. These may be helpful to improve the prediction of the trawl board motion before hooking occurs. Meanwhile, a parametric study on the hooking event could be carried out. Important factors in a hooking event such as span height, span shoulder slope, trawl board velocity, trawl board shape and crossing angle can be studied. Since the occurrence of a hooking event is quite sensitive to the trawl board motion, a systematic method of analyzing the hooking probability in real life condition should also be established. ACKNOWLEDGMENT The authors acknowledge Statoil for the contribution of trawl board hydrodynamic coefficients and model test results. The license of the software SIMLA is granted by MARINTEK and is hereby also be acknowledged. [9] DNV, 21. Offshore Standard DNV-OS-F11, Submarine Pipeline Systems. Det Norske Veritas, Høvik, Norway. [1] Sævik, S., 21. SIMLA - Theory Manual. MARINTEK. [11] Teigen, P., Ilstad, H., Levold, E., and Hansen, K., 29. Hydrodynamical aspects of pipeline overtrawling. In Proceedings of the 19th (29) International Offshore and Polar Engineering Conference, Vol. 2, ISOPE, Cupertino, CA, United States, pp [12] DNV, 211. Recommended Practice DNV-RP-H13, Modelling and Analysis of Marine Operations. Det Norske Veritas, Høvik, Norway. [13] Longva, V., and Sævik, S., 212. A penalty-based body-pipeline contact element for simulation of pull-over events. In Proceedings of the 32th (212) International Conference on Ocean, Offshore and Arctic Engineering, Vol. 4, OMAE, Rio de Janeiro, Brazil. [14] Nygaard, I., 22. Ormen Lange Gas Pipeline Overtrawling Study. Tech. rep , MARINTEK, Trondheim, Norway. REFERENCES [1] DNV, 21. Recommended Practice DNV-RP-F111, Interference Between Trawl Gear and Pipelines. Det Norske Veritas, Høvik, Norway. [2] Kjeldsen, S., and Moshagen, H., Influence of bottom Trawl Gear on Submarine Pipelines. Extension III. A Summary of Laboratory and Field Investigations. Tech. rep. stf6 f7912, SINTEF, Trondheim, Norway. [3] Moshagen, H., and Kjeldsen, S. P., 198. Fishing gear loads and effects on submarine pipelines. In Proceedings of the Twelfth Annual Offshore Technology Conference, Vol. 2, Offshore Technology Conference, Houston, Richardson, TX, United States, pp MS. [4] Nygaard, I., 199. Trawl - Pipeline Span Interaction. Model Tests. Final Report. Tech. rep /2, MARINTEK, Trondheim, Norway. [5] Verley, R. L. P., Moshagen, B. H., Moholdt, N. C., and Nygaard, I., Trawl forces on free-spanning pipelines. International Journal of Offshore and Polar Engineering, 2(1), May, pp [6] Verley, R., Pipeline on a flat seabed subjected to trawling or other limited duration point loads. In Proceedings of International Offshore and Polar Engineering Conference, Vol., ISOPE, Osaka, Japan, pp [7] Guijt, J., and Horenberg, J., Recent investigations concerning the effect of bottom trawl gear crossings on submarine pipeline integrity. In Offshore Tech. Conf, Vol., OTC [8] MAIB, Report of the Inspector s Inquiry into the loss of the Fishing Vessel WESTHAVEN AH 19 with four lives on 1 March Marine accident report 4/98, Marine Accident Investigation Branch, London, U.K. 9 Copyright 213 by ASME

10 Appendix A: Snap Shot Series of Both Simulation Visualization And Model Test for Pull-over Cases (a) SIMULATION FIGURE 14. (b) MODEL TEST PULL-OVER SNAP SHOT SERIES, H SP =.5m, α = 9 (a) SIMULATION FIGURE 15. (b) MODEL TEST PULL-OVER SNAP SHOT SERIES, H SP = 1.m, α = 9 1 Copyright 213 by ASME

11 (a) SIMULATION FIGURE 16. (b) MODEL TEST PULL-OVER SNAP SHOT SERIES, H SP = 5.m, α = 9 11 Copyright 213 by ASME

12 Appendix B: Snap Shot Series of Both Simulation Visualization And Model Test for Hooking Cases (a) SIMULATION FIGURE 17. (b) MODEL TEST HOOKING SNAP SHOT SERIES, H SP =.5m, α = 9, FLAT 12 Copyright 213 by ASME

13 (a) SIMULATION (b) MODEL TEST FIGURE 18. HOOKING SNAP SHOT SERIES, H SP = 5.m, α = 2, WITH SPAN SHOULDER 13 Copyright 213 by ASME