Using Field Data and Parametric Studies to Create a Dynamic Model of a Seismic Spread
Master thesis
Permanent lenke
http://hdl.handle.net/11250/2614965Utgivelsesdato
2018Metadata
Vis full innførselSamlinger
- Institutt for marin teknikk [3437]
Sammendrag
In a seismic survey, many demanding operations are done using expensive equipment that can be challenging to handle. A dynamic model may be useful to predict the response of a seismic spread for different scenarios with varying current and waves. This thesis researches a dynamic model of a seismic spread during a turn, including waves and current.
First, a literature review is conducted to collect information on numerical analysis of marine cables. The literature presents different methods for solving the static and dynamic equilibrium equations relevant for towed cables. Continuous and discrete methods are discussed and compared. The discrete methods are less accurate than the continuous methods, but demands less computations. Further problems mentioned in the literature are mainly for moored lines, and will not be a problem for a seismic spread. Literature including description of hydrodynamic and structural parameters relevant for towed marine cables is presented, to give a understanding of the possible range of these parameters.
OrcaFlex, a software widely used for modeling and analyzing marine systems, is used in this thesis to model the seismic spread. Models in OrcaFlex are used for two studies: case studies and parametric studies.
The case studies compare the model with field measurements from a real survey. The effective tension in the front end of the wide tow ropes is compared, since this is where the highest tension is present. Different elements of the model are found from field data, activity day reports and the project plan. The differences in model data and field data are linked up to different physical errors, and attempts to improve the errors are made. This comparison has been done for two separate cases, to verify the findings. The changes improved the average effective tension from around 40 % to 8-16 %. The differences in maximum and minimum tension have not improved, which is believed to be due to wrong wave loading. Further work should therefore include a study of wave loads.
The parametric study investigated how changing normal drag coefficient and bending stiffness affect the streamer behaviour. In this study, the position of the streamer ends for the different simulations is used to visualize and compare the behaviour. The range of the two parameters is discussed, based on theory from relevant literature, and simple calculations and different parameters within this range are tested. The normal drag coefficient gives big differences in behaviour. The simulations done in this work is for constant normal drag coefficient in the relevant range, assumed to be 1.2 - 2.4. The difference in track between normal drag coefficient 1.2 and 2.4 is about 200 meters at most. The normal drag coefficient for a real cylinder will be changing for the cross flow, and be 1.2 for most of the survey. Therefor the streamer is believed to behave close to the dynamic model with a constant normal drag coefficient of 1.2. To verify the conclusions a case study comparing positions from the model and from a real survey can be useful. This can also identify other parameters of interest. The difference in the behaviour for bending stiffness is about 1-2 meters at the most. This is assumed to be negligible, and bending stiffness can be any value between 50 - 50 000 kNm^2.