As a result of the steadily increasing activities related to marine technology in Arctic regions, Det Norske Veritas has launched an ice load monitoring project to gather knowledge of the ice conditions and prevailing ice-induced actions in the region. The intention of the following thesis is to study different aspects related to design of ice-going vessels, in particular the design scenario where a vessel impacts an ice ridge.The introductory part of the thesis gives an overview of important aspects related to sea ice, including different types of ice features and their physical and mechanical properties. The microstructure of pure ice and formation mechanisms of sea ice are briefly described, and mechanical properties such as elasticity and compressive strength are discussed. Further, a study of existing models for estimation of ice-induced loading on ships is carried out, with focus on local hull plating pressure and global loading due to ice ridge impact.A comparative study of design rules developed by Det Norske Veritas and the International Association of Classification Societies is conducted, and important differences between the two separate rules are identified. The subdivision of class notations is described, and differences in definition of design loads and corresponding requirements are presented. A general conclusion is that the rules developed by Det Norske Veritas are more specific when it comes to governing design scenarios, while rules set forth by the International Association of Classification Societies are more universal in terms of vessel type and prevailing ice conditions.Two separate finite element models based on coastguard vessel KV Svalbard are developed, including a simplified beam element model and a detailed shell element model. Quasi-static and dynamic response analyses for ice ridge impact loading are carried out, where the duration of the load pulse is varied systematically from 0.25~s to 2.0~s. The simplified finite element model is seen to give larger overall maximum response compared with the detailed model, but the difference decreases as the pulse duration is increased.It is observed that quasi-static response is overall larger than dynamic response for both finite element models within the defined pulse duration range. However, the ratio of maximum dynamic to maximum quasi-static response is seen to be positively correlated with the load pulse duration, and a close-to-linear relationship is observed.A study of different parameter variations is performed in order to investigate the importance of various pulse shapes, mass models, damping models and solution methods. Variations are only performed using the simplified beam model. It can be concluded that the shape of the load pulse is of minor importance for dynamic response when the pulse duration is short. However, the pulse shape becomes increasingly important for longer load pulses.An opposite trend is observed when varying the mass model, where a negligible difference in dynamic response is seen for longer load pulses. The difference increases somewhat for shorter load pulses, but can be considered unimportant for dynamic response within the investigated duration interval.It is further observed that the choice of damping model is of significant importance compared with other investigated parameters, and the difference in predicted response remains constant within the investigated pulse duration interval. The choice of solution method is however unimportant for analysis using the simplified beam model.In order to verify the applicability of the finite element models, full scale sea trial measurements of global motions from KV Svalbard are analysed and compared with finite element results. Difference between measured and calculated response during ice ridge impact is seen to be significant, where the calculated maximum response is close to 4 times larger than the maximum measured response. Iterative modifications of the load pulse shape are performed in order to reproduce the measured response history following ice ridge impact, and quite strong agreement is obtained between measured and calculated response.

SummaryThis Master’s Thesis goal is to present fundamental physical aspects of Vortex Induced Vibrations (VIV) of marine risers, and outline methods for suppression of VIV. Main emphasis has been given to the use of strakes, and relevant theories connected to strakes’ influence on excitation of riser, riser response and drag is presented. Variation of outer diameter of the riser is also studied.The theory has been put to test through case studies of two Aker Solutions in-marine workover systems of 321 and 1300 m water depth. The computer program VIVANA has been used. Several analyses have been performed for various different riser configurations. For the 321 m water system, the following configurations have been studied:-Base configuration, i.e. no use of strakes-Staggered configuration, staggered bare and buoyant joints-Bottom strakes configuration, bottom section of riser covered with strakes-Middle strakes configuration, middle section of riser covered with strakes-Top strakes configuration, top section of riser covered with strakesFor the 1300 m water depth system, the following configurations have been evaluated:-Base configuration, i.e. no use of strakes-Staggered configuration, staggered bare and buoyant joints-Middle strakes configuration, middle section of riser covered with strakes-Top 50_150 strakes configuration, top 150 m of riser bare, then coverage of strakes-Top stakes configuration, top section of riser covered with strakesFor each water depth a total of 4 different current profiles have been applied. The current profiles include both measured current profiles from the relevant oil fields, as well as several other more “theoretical” current models.The main findings from the 321 m analyses were:-Staggered configuration gives generally lower VIV amplitudes of the dominating frequency compared to base configuration.-Staggered configuration gives generally lower maximum stress amplitudes compared to base configuration.-No clear trends when comparing fatigue life of staggered and base configuration are found, however significantly better fatigue results found for the staggered configuration in measured current.-Maximum accumulated damage is located at the WH/XMT interface.-Top and middle strakes configurations give best VIV suppression results.-Applying strakes to the top section of the riser gives significantly lower VIV amplitudes, stress amplitudes and higher fatigue lives across all current profiles.-Top strakes configuration supress VIV completely for the sheared current profile.-Very small riser deflections and corresponding low flex joint angles are found; thus no operational consequences for the 321 m water depth. The main findings from the 1300 m analyses were:-A significant increase in active response frequencies compared to the 321 m water depth, more than 30 active frequencies calculated.-No clear trends in VIV amplitudes of the dominating frequency when comparing staggered and base configuration.-Highest stress amplitudes found for the base configuration in all current profiles.-No clear trends in calculated fatigue life when comparing the staggered and base configurations. Maximum accumulated damage found in the WH/BOP interface.-Top 50_150 strakes configuration the most efficient in suppressing VIV.-Significantly lower VIV amplitudes of the dominating response frequency for the top 50_150 strakes configuration. Same result found for the maximum stress amplitudes.-Compared to the base configuration significantly better fatigue lives found for the top 50_150 strakes configuration, however for the measured current profile an increase of only 1 decade was found.-Staggered configuration gives lowest static riser deflection for all current profiles, also after drag amplification from VIV.-Percentage increase in riser deflection from VIV reduced by roughly 80% when comparing the top 50_150 base configuration to the base configuration.-LFJ angles exceeded lower limits for certain drilling and workover operations, however applying the top 50_150 strakes configuration will generally give a larger operational window compared to the base configuration.Some of the results from the 1300 m analyses revealed certain discrepancies linked to the dominating frequencies and frequencies inducing maximum stress amplitudes. These inconsistencies are probably related to the convergence limit given as input in the VIVANA module.

Deployment of a large box structure in many sea states has been investigated. Two deployment methods are compared; crane installation over the side and through moonpool installation. The structure is 12 [m] long, 6 [m] wide and 12 [m] high with a mass of 250 [t]. Normand Subsea is used as installation vessel. Both JONSWAP and Torsethaugen wave spectra are used for crane lowering while only JONSWAP is used for moonpool installation.Splash zone lowering is seen as the most critical stage of the installation because hydrodynamic forces are largest at the surface. Hydrodynamic uplift is assumed limiting for the deployment, i.e “slack slings”. Slings are the lower part of the lift rigging. The operation limit is that dynamic uplift should not exceed 90 % of the modules static weight. Forces in z-direction are hence most interesting. Minimum wire tension for the lowering is therefore calculated at two time instances; when the module bottom end is at mean water level and when the module is fully submerged with its top end 0.5 [m] below mean water level. These time instances are referred to as time instance 1 and 2 in the report respectively. Design significant wave heights, Hs, are taken from plots of minimum wire tension for different wave peak periods Tp and wave headings. Based on these design Hs values, which equals the operation limits, operability rosettes are plotted. It is seen that wave headings 90⁰ and 120⁰ are most critical with lowest operability for crane installation while wave heading 90⁰ is worst for moonpool deployment.The lowest design Hs for all Tp values considering wave headings 0⁰ ±30⁰ is used as overall operation limit for deployment when weather window statistics are computed. Time instance 1 is worst for crane deployment with resulting forecasted weather operational criterion Hs=2.8 [m]. Time instance 2 is worst for moonpool deployment with forecasted operational criterion Hs=2.5 [m] and Tp ≥ 13.0 [s]. Reference time for deployment, hence the time needed from the weather forecast is issued to the module is landed on the seabed, is 6 hours. Based on reference time and forecasted operation limits weather window statistics are estimated.Moonpool deployment annual operability is 7.24 days, hence 2.0 % of the year, and can naturally not be used. Crane deployment annual operability is 213.22 days, 58.4 % of the year. This is much better but still not very good as it is theoretically desirable to be able to install the module any day year round.

Well intervention operations can inflict large strains on a wellhead. The seabed in the North Sea is very rigid up to the point where the sea water meets the mud or sand. This is one of the reasons why wellheads operating on the Norwegian continental shelf are more exposed to fatigue damage. In search for more oil and the wish to increase the utilization rate of existing wells the oil companies drill more and more, causing more and more fatigue life consumption. The oil companies must provide sufficient documentation that the wellhead always has enough remaining fatigue life to perform a Plug and Abandon (P&A) operation.Full scale measurements has been collected for a marine drilling riser connected to a moored Aker H-3 rig operating on a field with a depth of 325 m. The angles at the bottom of the riser above the lower marine riser package are used to calculate the consumed cumulative fatigue life using rainflow cycle counting and Miner-Palmgren summation.A simulation model has been developed in the computer simulation program RIFLEX, which is a state of the art simulation program developed especially for slender structures such as a riser in a marine environment. The model was built with relatively conservative assumptions. This resulted on fatigue life assessments that gave a shorter operation life than what was found using the full scale measurements. Using such simulation is often the only tool available to document fatigue life consumption since full scale measurement tools are rarely installed and used. It is vital that the simulation yield reliable and correct results and as close to the true result as possible.A series of similar simulation models were developed where we looked at the effect of taking away some of the conservatism in the original model. First we looked at the difference between a JONSWAP wave spectrum and a Torsethaugen wave spectrum. The difference lies in the assumption of that a sea state is a superposition of wind driven waves and swell waves, where the Torsethaugen use empirical data collected from the North Sea to account for the difference. A Torsethaugen is a double-peaked spectrum while the JONSWAP spectrum is a single-peaked spectrum. The difference between the two results gave little or no effect on the motion characteristics and fatigue life.Then we introduce a directional wave spectrum, meaning that waves may be short-crested and spread around a mean wave direction. This reduced the angular motion in terms of standard deviation significantly. The reduction was between 10% - 15%. It also affected the fatigue life positively. In the next model we introduced non-linear behaviour in the lower flex joint while the waves now were unidirectional. In terms of standard deviation the reduction was the same as for the model with wave spreading.In the last comparison model we used both non-linear flex joint behaviour and wave spreading. The total reduction was again significant. For some of the simulation even up to 30% compared to the original model. All the standard deviation from the full scale data has natural variances from data set to data set and most of the computer simulation fell within this margin of error.For all simulation models the model was tested with different mean heading direction of the waves. The mean heading directions were 0 deg, 30 deg, 60 deg and 90 deg relative to the rig. While the full scale measurement had little correlation between the measured response direction and the weather direction, the simulation were very consistent on this matter.Some simulations with current and support vessel offset was performed to find the effect on the standard deviation. While the presence of current damped the angular motion, the standard deviation increased with increasing support vessel offset. A discussion around the uncertainty of the true characteristics of the non-linear model explains some of the behaviour.When comparing fatigue life the calculated fatigue life consumption became closer and closer to the measured value as we removed the conservatism. However, by a closer investigation of the angle range spectra which is used in the Miner-Palmgren summation there was found differences that need more attention. While the angle range spectrum from the full scale measurement show a close to linear relation between numbers of cycles exceeding ranges the shape for the simulated models were far from linear. In terms of the shape parameter in the Weibull distribution it was found through fitting the curve that the shape model for full scale and RIFLEX simulations were around 1.05 and 1.9, respectively.It is this difference in shape that demands a closer investigation of the simulation models. The fact that the fatigue life approached the true fatigue life so closely should so far be regarded as a coincidence and not a result of good model approximations.It was also found that the full scale motion for some of the time series are low frequency dominated, i.e. high energy in oscillating components with a frequency outside the wave spectrum. Some peak periods reach periods over a minute or even two. This is an effect that is unaccounted for in the models presented in this thesis.

The age of the easy accessible hydrocarbon goes towards the end and new more challenging fields is the future of the oil industry. This is often synonymous with large depths where sea bed mounted platforms are highly uneconomical options. Floating platforms are therefore commonly chosen solutions.In this thesis two types of deep foating platforms are investigated; The spar buoy, a well known, well documented concept and the Octabuoy, a newly developed concept. Both platforms are analysed in two DNV programs, Wasim and Waqum. Wasim is a non linear time domain hydrodynamic program and Waqum is an impulse response function operator with the possibility of adding non linear eects. The spar bouy concept is used as a pilot test in the softwares. A recreation of the experimental results from H.A. Haslum's doctoral thesis from 2000 is attempted. The impact of non linear effects and mooring on both platforms is discussed. The subject of viscous damping is also approached.As has previously been confirmed by many researchers, the spar buoy is susceptible to non linear eects. The triggering of the Mathieu effect is shown in the Wasim analyses. Discussion is also made as to whether the spar might also be susceptible to large excitations due to second order difference frequency interactions between surface waves and body motions. Both these effects happen at low frequencies where radiation damping is low. Viscous damping is therefore of importance. From previous research mooring is found to be important to avoid the Mathieu effect by increasing the damping and moving pitch periods out of the danger zones.After analysing the spar buoy, the Octabuoy's motion characteristics are tested in mild to severe sea states in both softwares. Non linear effects are found to be significant in the vertical rotational degrees of freedom. The heave motion however seems relatively unaffected by non linear effects. Since Wasim models the free surface linearly, what makes the pitch/roll motion affected by non linearities is found to be either non linear hydro statics or non linear Froude-Krylov forces. The Octabuoy is designed to avoid the variation on hydrostatic stiffness. However, the deadrise angle is 10 degrees larger than the ideal angle. Whether this is what leads to non linear pitch/roll motion is not known at this stage.Two softwares are used in the thesis. Wasim has very long CPU time but calculates accurately and detailed information is easily accessed with for instance the Wasim application ForceInspector. Waqum is very quick, with CPU time in the order of minutes. The program requires an experienced user who knows what must be included for a complete analysis. There are uncertainties about the results from Waqum analysis and more verication is needed for the author to feel confident about the software.To conclude, the programs might work well together. Much can be tested quickly in Waqum, and then final configurations can be run more thoroughly in Wasim. It is the experience of the author that at least until the new version of HydroD is finished, running time domain analysis in Wasim should be done through scripting. This gives a larger control over the actual input and might decrease the chance for error.Keywords: Octabuoy, Spar Buoy, Non Linear Effects, Mooring, Wasim, Waqum.

Due to high expectancy of hydrocarbon resources in the Beaufort Sea itis seen as an important area for energy in the future. As the focus nowis shifting towards the deeper parts of the sea, there is a need for floatingproduction platforms that are able to operate year-round in the harsh con-ditions of the Beaufort Sea. In this Master’s thesis the design of such aplatform is performed, with respect to global ice loads acting on the plat-form and the capacity of the stationkeeping system. This is used to analysethe operability of the platform.To be able to develop a floating production platform it is crucial to haveinformation on ice conditions in the Beaufort Sea. Literature has been usedto determine ice conditions in the seasonal transitional zone, where it isexpected that the platform will operate. Factors such as level ice, first-year and multi-year ridge dimensions, floe size and ice island mass havebeen defined. These define some of the most common ice conditions in theBeaufort Sea, and as such represent ice conditions a platform can expect tomeet.Functional specifications have been used to determine the most appropriateplatform type. Several platform types were considered based on factorssuch as production and storage, ice loads, and ice interaction with risersand mooring lines. A circular FPSO was selected due to its insensitivity toice drift direction. To reduce ice loading, the platform was designed as adownward sloping cone.Ice loads on the platform have been calculated using several methods com-monly applied in literature. Results show that managed ice yields lowestloads, followed by level ice. Higher loads are seen for ridges, both first-yearand multi-year, and large ice floes.A water depth of 500 m was assumed for the location of the platform. Since this can be considered as a moderate water depth, a conventional mooring system was selected for stationkeeping. The system has 24 mooring lines connected to a buoy, which again connects to the turret on the platform. Internal mooring was seen as necessary to protect mooring lines from ice. Restoring coefficients have been determined using the inelastic catenary equations. Maximum allowable horizontal displacement is defined as 5 % of the water depth. This is used to determine the maximum allowablehorizontal load on the platform to be 53.38 MN.Lastly, the operability of the platform was judged by comparing various iceloads to the maximum allowable load given by the mooring system. Icemanagement is clearly needed when operating in any ice other than levelice. The requirements of the ice management system has been defined, anda general analysis of the primary and secondary objectives of the ice man-agement system has been performed. It is proposed to use one icebreakerthroughout the year, since the platform may encounter multi-year ice floesduring the summer. As the ice season starts an additional ice breaker isintroduced. In severe ice conditions it may be necessary with a third icebreaker. It is concluded that if proper ice management is ensured, the plat-form should be able to operate year-round.

Non-linear analysis of wind load subjected flare tower in Abaqus

Computational Fluid Dynamics (CFD) have been applied for numerical simulations of the flow around simplified shapes of a Remotely Operated Vehicle (ROV). The simulations have been carried out at Re = 850 000, which coincides with a free stream velocity of one knot in seawater conditions. Detached Eddy Simulation (DES) have been used with the k-w SST turbulence model in the near-wall (unsteady RANS) regions. The CFD code FLUENT from ANSYS Inc. have been used.The first case to be investigated, was a fully submerged rectangular stationary cube with both sharp and rounded edges. The sharp edged case showed good agreement with published reference values. The results for the refined mesh gave Cd = 0.8192, Clrms = 0.0473 and St = 0.1132. The Strouhal number was identical for both meshes, while small discrepancies were seen on the mean drag and the root mean square lift coefficient. For the round edged case, the mean drag coefficient was shown to be about 25% of the sharp edged case (Cd = 0.2257). The lift was more difficult to model correct and large discrepancies were seen both on the Strouhal number and the root mean square lift coefficient. It was concluded that longer time series were needed in addition to further mesh refinement in order to get more stable mean quantities of the lift history. Some effort were also made on trying the realizable k-e turbulence model in the near-wall regions for the round edged case, but without noticeably effect on the results. For both the sharp edged case and the round edged case it was concluded that although the drag was successfully modeled, a further mesh refinement was needed in order to ensure converged results.For the case of the simplified model of the ROV, three meshes were used for the grid convergence study. Small discrepancies were seen, between 2% and 6.2%. The pitch moment through the Center of Gravity (COG) was also measured and a negative mean value of Cm = -0.2563 was obtained. This means that the reported instability for the ROV was also seen in the simplified model. Through a more detailed study of the pressure and velocity distributions, the main problem regions were identified to be above the forward top and below the aft bottom. Here high-velocity regions generate low-pressure regions, which coincides with a negative pitch moment. One proposed solution was the removal of the plate in the aft which span the width of the ROV. The effect of this solution was however found to be questionable. This is due to the components which are placed in front of the plate, and these may actually cause the same effect. For this reason, these components should be relocated.It was concluded that there should be performed numerical simulations without the plate and the components in front included. The effect of these components could then be studied and a decision on whether to remove the plate could be made. Also, numerical simulations where different locations for the components in front of the plate are tried should be an interesting case to run with respect to the stability analysis.

The World’s energy demand is rapidly increasing and a good viable renewable energy source is wind power. The land-based knowledge and experience the onshore wind turbine industry possess is used to develop offshore wind turbines. With this knowledge together with the experience and knowledge of the marine industry we can design and produce a floating wind turbine. The main advantages of an offshore wind turbine are that the wind is stronger and less turbulent at sea, visual and noise annoyances can be avoided and there are large available areas at sea. In this thesis coupled time domain analyses of a floating spar-type wind turbine are performed with the intension to study parameters affecting fatigue damage at base of the tower. The software applied is SIMO/Riflex with the extension TDHmill, which gives the wind thrust force and gyro moment on the wind turbine as point loads in the tower top. Short term environment conditions are chosen from a joint distribution of simultaneous wind and waves which is based on measurements from a site in the North Sea in the period 1973 – 1999. In total 141 different environmental conditions are chosen for the sensitivity study. Mean value, standard deviation, skewness and kurtosis are calculated for axial stresses at the base of the tower. Fatigue damage is calculated from the Palmgren-Miner sum with a nominal stress SN-curve from the DNV fatigue standard. The axial stress-cycle distribution used in the Palmgren-Miner sum is found by rainflow counting. Time domain simulations are carried out for the different sea states and fatigue damage is calculated for each case. The statistical properties and fatigue damage are averaged over seven samples with different random seed number to ensure acceptable statistical uncertainty. Accumulated standard deviation shows that 5 samples of each load case are sufficient to ensure acceptable statistical uncertainty. Sensitivity study of different simulation length shows that 30 minute simulations give close to equal fatigue damage and standard deviation as 2.5 hour simulations. Sensitivity of fatigue to wave height and peak period is carried out to study the effect of varying parameters. This study suggests that the highest waves dominate the fatigue damage for the smallest peak periods. For some small wave heights the damage will be constant for a given peak period range. From this sensitivity study it is shown that if the deviation of fatigue damage between the different load cases is small, then the dominating load case of the accumulated long term fatigue damage will be dominated by the marginal probability of each load case.

Present regulations for offshore structures on the Norwegian continental shelf havea requirement for Accidental Limit State redundancy and Fatigue Limit State redundancyin a damaged state. However, the requirement is far more defined forthe Accidental Limit State than for the Fatigue Limit State. An increased understandingof factors governing fatigue redundancy would create a basis to form acomprehensive definition.In literature, the term redundancy is defined in several ways. The different methodscan be divided into two major categories, namely deterministic and probabilisticapproaches. In general, redundancy may be defined as the absence of memberswhose failure would lead to global collapse. Within both the deterministic and theprobabilistic framework, several redundancy factors are usually defined, and thereare resemblance between some of them. In probabilistic methods, the reliabilitymethod is commonly applied through the First Order Reliability Method.Fatigue damage is a primarily concern regarding the integrity for offshore structures.A near constant subjection to cyclic loadings from wind, current and wavesinitiates a cumulative damage process which leads to a certain fatigue life for themembers in the structure. The lifetime may be calculated using either a fracturemechanics approach or a SN-curve approach. There exists several approaches tocalculate the stress levels to be used in the fatigue analyses, and the choice ofmethod is mainly based on whether or not the structure under consideration isdynamically behaving or may be regarded as quasi-static. Also, there may be nonlinearitiesthat needs to be accounted for and naturally this will affect the choiceof analysis method. However, large uncertainties are associated with fatigue calculationsregardless of analysis approach. Thus, a probabilistic framework is highlyrelevant in order to estimate the risk of failure due to fatigue.A study on how impaired integrity affects the fatigue life has been performed fortwo jackets; one highly redundant four-legged jacket and one less redundant threeleggedjacket. The main goal has been to investigate the fatigue redundancy of thestructures, in order to link up the risk of accelerated fatigue due to damage withthe risk of failure due to extreme environmental actions, which may eventually leadto structural collapse. A stochastic fatigue analysis approach was chosen, and theanalyses was in agreement with the standards governing the Norwegian continentalshelf at the time of this thesis. Pushover analyses has been performed on the jacketsto give an insight in their redundancy, and a calculation of changes in the naturalperiods under impaired integrity has also been done.The four-legged jacket was proven to be highly redundant, and had small changesin the natural period under impaired integrity. The three-legged jacket on the otherhand, had some damage cases with a rather large increase. Also, the redundancyfactor R4 was significantly lower for the three-legged jacket, thus confirming it tobe less redundant than the four-legged jacket. A large change in the natural periodwill alter the dynamic response, thus the fatigue life is vastly connected to changesin global stiffness.Trying to explain the changes in fatigue life for the two jackets under impairedintegrity without using a deterministic approach, i.e. calculate the fatigue life forthe specific damage case, was proven to be very difficult. There seems to be noeasy way to isolate the severity of the fatigue life reduction since large changesare occurring throughout almost the entire structure for several damage cases, aswell as large spread in the values them self. However, there was also found sometrends in the results. One of them, was that the closer a member is to the damagedelement, the larger is the expected reduction in fatigue life. Another trend, is thata large fatigue accelerator factor is expected in almost every damage case, thus onemay expect large changes for most of the damage scenarios.Another vastly occurring phenomenon were the location of the damaged memberswho gave the lowest fatigue life in the structures. For the four-legged jacket, thisinvolved damage in the caisson supports. The three-legged jacket, however, hadthe lowest fatigue lives occurring for damage cases in the top frame where there isa lack of deterministic redundancy.There seems to be no fatigue redundancy for the jackets, as there are large fatigueaccelerator factors occurring. There is also the very low calculated fatigue life inthe most extreme cases. However, there has been found a slight correlation betweena large reduction in fatigue life and a large initial fatigue life. What should also betaken into account though, is both the risk related to the fatigue lives found, andthe accuracy of the values due to the linearised analysis.