Ships transporting different liquids have been around for as long as there has been the need to trade between countries and nations. The first liquid transport happened in drums loaded into the ships like any cargo. During the last 120 years as ships have been built of steel/iron it has been possible to transport liquids in tanks built in to the structure of the ships. During the 20 th century cargo ships sailing in ice were assisted seasonally mostly by icebreakers. Most of the liquid transport did not really need any assistance in ice conditions as there was hardly any winter traffic. The continuous need for utilizing oil lead to new territories and the oil discovered in Alaska in the late 1960ies brought up the crude oil transport in ice conditions and Esso started the Manhattan project. This unfortunately did not start the arctic tanker transport. This paper discusses the development history of the propulsion of tankers working in ice conditions. In the early 1970ies the winter traffic in Finland exploded as the Finnish government decided to keep all the major ports open year-round. This started serious development of ships capable of more independent ice operation. The first propulsion solutions were modifications of well proven technology. We had fixed pitch and controllable pitch propellers connected directly to low speed diesel engines or through a gearbox. In some applications the propeller was equipped with a nozzle. In the 1970ies and 1980ies we were already drafting different tankers with electric propulsion with huge power. However, it was needed a decade more and the development of the propulsion devices from direct shaft lines to azimuth devices to get the propulsion revolution going at full speed. Today we have available and most commonly used; traditional fixed pitch propellers with conventional shaft lines, mechanical Z-drives and podded propulsors driven by electric motors. The operational profile and mission of the vessel will dictate how the tanker will be furbished. Most of the new concepts have been tested and developed in conjunction of icebreaker development, where tankers have been closely following. Recently there have been delivered and designed new icebreakers, icebreaking shuttle tankers and LNG carriers. Many of these vessel concepts are relying on podded propulsion system. AZIPOD propulsion has been selected to many of these vessels as it provides excellent ice performance for the vessel, good torque characteristics for the propeller and there already exists proven track record of ice operations. This paper will introduce some outstanding full-scale experience from the revolutionary icebreaking LNG carrier Christopher de Margerie.

This paper presents operational experience and ice measurement results from Finnish Icebreaker Polaris propulsion system. Icebreaker Polaris applies a novel three propulsion unit concept. During the shipbuilding period an extensive ice load measurement system was installed on two of the Azipod units. The paper will present key findings from the measurements from the winter periods 2017-2018. The target of the ice load measurement campaign is to increase the designer knowledge of propulsor ice loads in different ship operations. The purpose is to understand in more detail the load distribution for component life estimations and clarify differences between bow and stern propulsor ice loads and frequencies. Due to non-linear load distribution in propulsor at extreme loads, strain gauges were installed at two levels in propulsor hull. This paper will highlight some measurement results of propulsion system behavior during the ice trials and long term measurements from the winters 2017 and 2018. The full-scale ice operation data collected in the Baltic Sea ice conditions can be further utilized in a development vessel and its propulsion system of heavy polar icebreaker type. The measurement campaign onboard the Polaris is planned to continue during coming winters. The aim is to collect statistical ice load data to improve the understanding of the fatigue and ultimate design loads. One of the goals is also to clarify how much load the bow mounted propulsion unit will experience compared to the stern units. Based on the measurement results so far, we can conclude that icebreaker with bow mounted Azipod propulsor can be operated without limitations in all ice conditions, just like all the other icebreakers in the Baltic.

This paper describes the results of model tests carried out in ice and open water conditions to evaluate performance of the United States Coast Guard (USCG) Heavy Polar Icebreaker (HPIB) indicative design. The resistance, propulsion and manoeuvring performance in ice conditions was evaluated at three different ice thicknesses (4, 6 and 8 ft.) with flexural strength 100 psi using two power setups, 36500 HP and 65000 HP. Calm water resistance and propulsion tests were also performed to evaluate open water performance. Models were constructed and tested corresponding to two indicative designs, one with triple shaft propulsion system and the other with one centre shaft and two wing podded propulsors. The present paper describes only the results for the model with triple shaft propulsion system.

The Polar Class Module Carriers AUDAX and PUGNAX were built for a unique service; year-round transportation of ultra-heavy construction modules for the Yamal LNG plant which is located near Sabetta on the Yamal Peninsula, 4 degrees north of the Arctic Circle. The average- and winter conditions in the Russian Arctic, especially the Ob Bay area, are extremely harsh. Dedicated Vessels were essential to provide the Client the required year-round module transportation service in these challenging conditions. The unique design of the sister vessels AUDAX and PUGNAX is jointly developed by Aker Arctic Technology and RED BOX Energy Services. A special custom/one-off hull shape has been engineered to ensure the Vessels can perform reliable module transportation in Arctic conditions within a limited timeframe. The requirement for stern-to loadout and discharge operations made a conventional hull shape or double acting ships (sailing a-stern in ice) design impossible. With the aid of CFD calculations and model testing at Aker Arctic's own test basin, the unique hull shape for the Vessels was developed. With the CFD analysis a new ice bow was designed specifying icebreaking capacities to 1.5 meters of level ice. Results of the performed tests demonstrated the Vessels design exceed the minimum required performance. Furthermore, in close cooperation with DNV-GL the Vessels obtained Polar Class 3 (PC-3) approval. In addition, two special studies were done by DNV-GL and the CNIIMF to proof compliance with ARC-7 Class, the Russian equivalence of PC-3. The bow design of the 43 meter (141 foot) wide Vessels enables them to break ice up to 1.5 meters (5 foot) thick, independently. AUDAX and PUGNAX can cope with temperatures of -50°C (-58°F) and still navigate safe and reliable during the 9-months ice coverage period in the Arctic, including the challenging polar nights which last for two months annually. Both Vessels were delivered early 2016 and have, since then, profoundly navigated around the clock; AUDAX and PUGNAX delivered 48 modules at the Yamal Peninsula while safe navigating 30.000 mile through ice covered waters. In March 2017, Aker Arctic and RED BOX jointly assessed the performance of AUDAX, when the Vessel sailed from the Yamal Peninsula to Zeebrugge (Belgium). The overall performance of the AUDAX was impressive, performing well-above its specifications in both ice breaking conditions as well as in modules loading- and discharge operations. AUDAX and PUGNAX are one of the largest icebreaking Vessels in the world. Combining these icebreaking capabilities with a unique module transportation approach, RED BOX created an extraordinary Logistic Service to serve large-scale Energy Infrastructure projects in Arctic Regions. The additional special features of the Vessels which includes deck heating and internal ballast transfer make efficient transportation and cargo operations in extreme harsh environments feasible.

Wind turbine towers are being planned in ice covered regions subject to pressure ridges (e.g. the Great Lakes). Conical collars are often employed to reduce ice loads from level ice and their associated dynamics. For level ice, downward breaking cones have some advantages. It is not clear if this is the case for pressure ridges. This paper presents an improved method for ridge loads on wind turbines with downward breaking cones and makes comparisons with upward breaking cones. First year pressure ridges can be formidable ice features and usually control design ice loads in the sub-Arctic. Important components of a ridge creating ice loads are the consolidated layer at the surface (which is considered as solid ice) and the ridge keel below consisting of ice rubble, but much thicker. The load due to the consolidated layer is usually derived as if it is thick level ice. On a cone, methods for level ice assume it can be idealized as a plate on an elastic foundation (the water) and equations have been developed for upward and downward breaking cones. But for a ridge on a downward cone, to break the consolidated layer downwards requires it to be pushed into the keel rubble below. This will have a different foundation modulus than water buoyancy. A method is developed to account for this difference. The method uses an iterative approach to determine the point of breaking of the consolidated layer (and associated load) accounting for the ridge geometry, keel rubble shear strength, the flexural strength of the consolidated layer and the buoyancy forces. The keel loads on the vertical shaft below the conical collar are based on the method currently in ISO 19906 (2010) but modified to add the effect of additional rubble in the keel from breaking the consolidated layer downwards. In examples, it is shown that the breaking force can be about twice that of breaking the consolidated layer without the keel present. This might be seen as a disadvantage for downward breaking cones vs upward breaking. However, it is also shown that the clearing forces on an upward cone are higher; which tends to balance out the lower breaking force. Example loads are given on typical wind turbine bases due to typical ridges. Upward and downward breaking configurations are compared. The paper provides new methods for ice loads due to ridges acting on wind turbine structures not currently covered by existing methods.

The paper addresses methods used to analyse full scale trials for performance determination of ships in ice. Furthermore uncertainties resulting from assumptions taken in the early stage when ice model tests are typically performed are discussed. In a first step the difference between model test based prediction and final determined performance from ice trial are shown for an example ship for which both results are available. The main assumptions are critically reviewed and possible reasons for deviations between the prediction and final outcome of ice trials are systematically assessed. The performance of ships in ice determined by model testing based on single combinations of speed and ice thickness is compared to full scale trial results. Furthermore the effect of generalizing impact friction and snow cover which is usually generalized for the model based prediction is determined by different methods. Finally a comparison between the propeller-ice interaction based on model tests and full scale trials is made to assess possible sources and magnitude of deviations between the performance prediction in early design stage and the final trial results. The paper shows that even though the prediction of icebreaking resistance by theoretic means and by model testing is well established comparisons between performance results obtained from ice trials sometimes deviate in their values and trends from those predictions. Besides testing uncertainties also missing parameters and testing methods can be identified as reasons for differences in the results. Recorded time series of propeller ice interaction in full scale and model scale are compared and analyzed. Different scaling approaches are applied to the model scale torque trends and compared to the full scale data. The results of the comparison between full scale and model test results show the limits of the current methods and encourage to take a closer look at additional parameters with stronger influence on ice performance. Additionally the comparisons of ship performance and propeller ice interaction of model and full scale data will improve the understanding of the transferability from model to full scale data.

Numerous field studies show that mature first-year (FY) ridges consist of a solid ice consolidated layer (CL) below which is a keel of ice rubble. The CL is generally thicker than the ambient level ice. With the decrease in multi-year ice in the Arctic, FY ridges may become the controlling ice feature at many Arctic locations. In sub-Arctic regions ridges are smaller but they also usually control design ice loads. For vertical structures, ISO 19906 (Standard on Arctic Offshore Structures) suggests calculating the load due to crushing of the CL and adding this to the load to fail the keel. In ISO a formula is provided for the load to fail the keel on a vertical-face which assumes that the keel material has Mohr-Coulomb properties. No specific algorithms are given in ISO for FY ridge loads on sloping structures. The work described in this paper addresses this gap. The study investigated three bounding methods and compared them with the measured loads due to FY ridges on Confederation Bridge (which has piers with upward breaking cones). The three bounding methods are: Model A; which assumes the CL breaks in bending and rides up as level ice and the keel load is calculated assuming a "dead wedge" is created on the slope which converts the slope into a vertical face against which the keel fails. This model can make use of the methods in ISO 19906 for calculating these two components and can be considered to be the implied "current ISO approach". Model B assumes that the CL layer fails in bending as "level ice" on an elastic foundation and rides up the slope with the accompaniment of additional ice rubble scooped-up from the sail and keel of the ridge. Model C assumes that the FY ridge can be idealized into an equivalent "solid ice" beam using composite beam theory. Then the beam on elastic foundation method, as used for solid ridges, is used to estimate breaking and clearing loads. These various approaches are reviewed and the derived loads are compared to failure modes and measured loads from Confederation Bridge for selected events involving FY ridges. Based on these comparisons a hybrid of Models B and C is recommended and the paper gives the details of how to apply this method. When used for example structures, the new model gives loads which are 40 – 50% lower than the current approach implied in ISO 19906. The method can be adapted to downward sloping structures.

Marine vessels and offshore structures functioning in Earth's frigid zones require ice management to continue their routine operations. Icebreakers are the most influential vessel in assisting marine operations in Polar Regions. The present study is set to analyze the clearance area of level ice using Azimuthing propeller jet in bollard condition, by means of full-scale and model scale experiments. Moreover channel widening and heeling test is performed to analyze the escorting ability of an icebreaker with only using propeller jets. Scope of the current investigation can be incorporated in designing new icebreakers and maintaining desired channel width based on propeller jets effect. Propeller jets can be used to break level ice, when the ship is stationary or moving, where the amount and capacity of breaking or clearing the ice are based on the thrust of the propeller, angle between the propeller jet axis and free surface, and thickness of the ice as well as propeller running time. This paper presents a comparison between full-scale experiments data (carry out in the Gulf of Bothnia, March 2017) and model scale trials performed in Aker Arctic testing facility on the level ice sheet. These experiments were based on image data from external camera and propeller flow parameters, where the area, as well as coordinate calculation, were within 3% of the accuracy from the acquired images. Full-scale ice thicknesses utilized in the experiments were selected and confirmed from surveillance videos. Model-scale images were corrected using Hugin software while ImageJ was used to calculate ice clearance parameters. Propeller thrust and area analysis show 10-22 % of the variation in the results of the model and full-scale experiments for 16 mm thick model ice. 16 mm thick model ice results are much closer to full-scale trials than 25 mm thick model ice. Test results at 30° and 90° pod angles could be extrapolated to design a prototype vessel. Channel widening shows worthy outcome, with the use of Azipods at a speed of 8 kn channel width of 36 m can be attain by positioning the stern Azipods at 30° puller configuration. Changing the pod inclination by 30% will increase the channel width to 20%. In the widening of new level ice channel, 30° pod angle is the most suitable.

Multi-year (MY) ridge and level ice interactions with sloping and conical structures involve complex ice feature shapes and ice failure mechanisms. The limited available field data makes calibration of associated load models difficult. To account for associated randomness and uncertainty, models may tend to be on the conservative side. New deterministic algorithms were recently developed to calculate loads more accurately for interactions of MY level ice and MY ridges with an upward sloping structure. This paper presents the application of these recently developed formulations in a probabilistic framework using SILS. SILS is a Monte-Carlo type simulator developed by C-CORE following the general procedures outlined in ISO 19906. Ice and metocean input parameters are defined by the user either as a fixed value (e.g. friction coefficients) or a random distribution (e.g. ice drift speed, floe size). The yearly encounter frequency is first estimated for these ice features for the site of interest. The ice loads are then determined for each of simulated interaction event occurring over the model timespan, using the deterministic load formulations. By simulating a large number of years of ice interactions, design ice loads can be determined that correspond to various low annual probability of exceedances. This paper demonstrates how complex loading scenarios, modelled in terms of idealized deterministic models, can be incorporated within a Monte-Carlo framework to provide design level loads. During the model implementation and analysis of results, significant improvements were identified and implemented in the deterministic model, resulting in a more robust model and better design estimates. The results provide valuable insights regarding model inputs and behaviour corresponding to the extreme design ice loads. An example of a full probabilistic analysis is presented in the paper to illustrate the models. Here the probabilistic framework of SILS is used to assess a Base Case scenario and a number of sensitivity cases using different environmental inputs and model parameters.

AKAC participated in the full scale ice trials of the newly built R/V Sikuliaq, owned by the National Science Foundation and operated by the University of Alaska-Fairbanks. The purpose of the trials was to identify its ability to conduct science missions in ice, as well as to develop operational procedures for conducting independent science missions in ice. This paper gives an overview of the key design features of the R/V Sikuliaq that enable it to conduct science missions in ice, as well as the observations and results from the full scale trials. This paper provides real life experience on the operational performance of a state of the art research vessel. The experiences shared in this paper are applicable to a wide range of operations in ice, including station keeping in ice and towing in ice.

Baltika, the world’s first oblique icebreaker designed to break ice sideways, left Murmansk on 20 March 2015 and headed to the Russian Arctic. The purpose of the three-week voyage to Kara Sea and the Gulf of Ob was to evaluate the vessel’s icebreaking performance and operational capability through extensive full-scale trials in challenging Arctic ice conditions. In the ice-free Barents Sea, the seakeeping characteristics of the asymmetric icebreaker hull were also evaluated in moderate seas. Performance trials were carried out in three different ice thicknesses, ranging from 40 cm thick saline sea ice in the Kara Sea to up to 1.22 m thick hard low-saline ice outside the Sabetta LNG terminal in the Gulf of Ob. During these trials, Baltika exceeded her design icebreaking capability of 3 knots in 1 m thick ice in both ahead and astern directions. In addition, the oblique icebreaking mode was demonstrated for the first time, and the vessel performed beyond expectations. While the main goal of the trial voyage was to confirm Baltika’s icebreaking capability, both the vessel’s crew and the designers gained considerable operational experience during the daily operations in the challenging ice conditions. After the best way to tackle obstacles such as ridged ice fields was discovered, Baltika was found out to be equivalent - sometimes even superior - to conventional icebreakers despite her lower propulsion power. There have always been those who have doubted the feasibility of the oblique icebreaker concept. The extensive full-scale ice trials in the Russian Arctic have shown that Baltika, the first icebreaker with an asymmetric hull, could not only break ice sideways, but also sometimes out-perform conventional icebreakers in other operational situations as well. The concept is thus seen to hold potential for a number of missions such as escort and port icebreaking and in offshore projects.

The hull form and propulsion are essential for a successful ship. During the first 70 years of last century vessels for ice operation were developed mainly with trial and error method. Also when one design was found useful, it was duplicated to the next generation of ships. It was only in the mid 1950ies as the first model basing for developing ships for ice operation in Leningrad, USSR, started to take steps towards more scientific approach. In the late 1960ies after oil had been found in Alaska and the 100000 DWT tanker SS Manhattan had done its first voyage in the Canadian Arctic in 1969, the question rose, whether this could be modeled in smaller scale. SS Manhattan was the actual initialization for developing ice model tests to become a useful tool in modern shipbuilding. Since 1969 the testing methods and practices have changed and developed together with feedback from full-scale ship performance tests to be rather reliable tool to help the decision making process when we are talking about projects worth tens and hundred million dollars. This paper talks about different aspects of icebreaking and steps taken to develop vessels using ice model tests. Both hull form and propulsion development are discussed.

The Sea Ice Loads Software (SILS) is a Monte-Carlo type simulator developed by C-CORE for determining first and multiyear design sea ice loads, following the ISO 19906 methodology. Due to complexities in ice failure mechanics and associated uncertainties, these models are necessarily empirical or require simplifying assumptions. To account for uncertainty, conservatism must be built into models. A number of improvements to the software have recently been implemented in order to more realistically include a number of model components applicable to multi-year ridge loads and limit forces. This paper provides an overview of new modules accounting for ridge breakout, driving force ramp-up and ridge geometry modeling. Sensitivity runs show that design loads are affected significantly by accounting for these processes, compared to the previous implementation of ISO 19906 in SILS. The objective of this paper is to present the analytical model components and their implementation in SILS, and to demonstrate the influence of the changes by means of a scenario wherein a vertical sided structure encounters multi-year level ice and ridges.

Abstract This paper presents the results of a study examining the feasibility of a Conical Piled Monopod (CPM) offshore Northern Alaska in about 41m water depth, fairly competent soil conditions, and Multi Year Ice environment. The study was conducted jointly by ConocoPhillips Company based in Houston, Texas and Granherne Ltd, a KBR Company based in Leatherhead, UK. ConocoPhillips sponsored the study. Two CPM concepts were considered, the Standalone CPM covered in this paper and a Dual Rig Jack Up (Gemini) Assisted CPM. The latter is the subject of another paper to be published later. The CPM was configured for two drilling rigs, assuming topsides weights ranging from 35,000 tonnes to 45,000 tonnes and corresponding operating weights from 60,000 tonnes to 70,000 tonnes respectively. Ice loads were calculated using ISO 19906 formulae where possible, otherwise using references within the ISO 19906 code and specialist ice load advice (Ken Croasdale) where necessary. The CPM has been shown to be feasible in the Alaskan Sea for the two ice conditions considered, 6m thick level ice and a very rare 25m thick ice island event. The ice island event represents a very conservative scenario and further data may allow its severity to be reduced. In addition, there will be many months, if not years, after discovery before an ice island can potentially impact the CPM. The CPM has a large enclosed volume available, which could accommodate a potentially significant amount of oil storage. However, this has not been quantified within this study work. The CPM concept eliminates or greatly minimizes the need for seabed preparation, being a pile supported structure. It also does not require solid ballast for resistance against sliding due to ice loads; these two characteristics of the CPM can have a major positive impact on the cost and schedule of an offshore Arctic project. ConocoPhillips has a Patent Pending on the CPM concept.

Abstract The increase of activities in oil and gas exploration and production in cold regions during the last years is associated with the development of new innovative concepts with respect to arctic fixed and moored offshore installations. As a consequence there is a huge demand from the oil industry to validate the proposed concepts and feasibility studies. Apart from that, Numerical simulations may be beneficial to a certain extent to investigate the ice related problems, one can still not dispense with the execution of physical model tests. Worldwide there are only a limited number of facilities available, which are able to cope with the challenge of ice model tests at a proper model scale. The paper describes broadly the assets and drawbacks of different types of mooring systems ("wet" or "dry" mooring system) and the choice of suitable substituted mechanical systems with respect to their application for various ice scenarios. Alternatives how to design the model test set-up in order to measure the ice forces independently from the system response forces of moored floating structures will also be discussed.

Abstract Today's best engineering practice for the design of a stationary floating structure in ice infested water is articulated around physical testing in an ice basin and numerical modelling of the structure response to the relevant ice interactions. Ice basin testing is the best way to simulate the complex scenario of a moored floating structure interacting with drifting intact level ice and ridges. It is also per today the main source of experience for the design of such structures as full scale experience is sparse. However, ice basin testing suffers from deviations due to scaling effects or mismatch between one or several of the achieved and targeted properties (of the modelled ice and structure). As a consequence, the outcome of ice basin tests needs to be assessed and correction of measurements are required. An efficient way is to use an engineering numerical model to simulate the achieved ice-structure interaction in the ice basin, qualify the numerical model by comparing simulation results and measurements, and then simulate the targeted interactions with the numerical model. Such a numerical model may further be used to simulate the full range target design conditions for the ice structure interaction. The RITAS ice basin campaign tested a structure element designed to gain insight in the level ice bending and accumulation process around a traditional moored floating arctic structure design. These measurements are well suited to assess the validity of a numerical model designed to replicate this type of interaction. The SimShipIce numerical model is an engineering tool which focuses on replicating the steering processes during the interaction between a moored structure and drifting intact level ice and ridges. Such a model needs to be calibrated against ice basin test outcome.

Abstract In the early 1990's Russia was looking for the possibility to build a new oil export channel from the Timan — Pechora area through a pipeline to a port located in the Gulf of Finland. Various port options were studied and the final export harbour location was decided to be in Primorsk, Russia in the Eastern end of the Gulf of Finland. From Primorsk the crude oil would be shipped to commercial markets by tankers Already at that time it was seen that the increased oil transportation in the Baltic Sea would mean larger size tankers. This led to new requirements for the of icebreaking operations to assist the vessels, which are wider than the assisting icebreakers. With the existing icebreaker fleet two icebreakers are needed to create wide enough channel for one tanker. To avoid using two icebreakers and reduce the operation cost, one vessel would need to be able to make both wide and narrow channel. The oblique icebreaker concept was developed in the mid 1990's by Aker Arctic (at that time Kvaerner Masa Yards Arctic Research Centre - MARC). The idea is to use one single vessel with two operational modes; normally bow / stern ahead or in an oblique angle. This would enable the vessel to create both narrow and wide channel. The oblique operation would mean that the ship hull is not symmetric. The general requirements in asymmetric ship design has forced to new thinking also in other areas, which in turn has re-opened the commonly used solutions and designs. The new unconventional concept has opened wide range out-of-the-box possibilities also in other areas, such as oil spill combatting, ice management in harbours and offshore. The novel concept has seen many stages of development and now it is becoming reality as Russian Ministry of Transportation decided to order the first oblique icebreaker in 2011 for Baltic operations. The first vessel (NB508 at Arctech shipyard) is under construction, when this paper is being written. The operation of the new concept will also be a learning process for the captains and will in time further reveal new possibilities for the vessel use. Experience from the design of the Baltic version is used to further develop the concept for Arctic use. The Arctic version was introduced to public in May 2013. This paper shows the technology behind the vessel under construction and also shows major principles of the concept and introduces the new possibilities of the oblique ship concept.

Abstract DYPIC – Dynamic Positioning in ICE - is an international research and development project partially financed by the national research agencies of Germany, Norway and France. This 3-years initiative (2010–2012) focused on the development of Dynamic Positioning (DP) technology for the Arctic environment. The projects' backbone was formed by two extensive experimental campaigns performed in 2011 and 2012. This paper summarizes the work performed within the project and spotlights the technical and scientific findings emerged from it. Special attention is payed to two facets of the project: the design of experimental devices, systems and setups for ice tank testing including the development of a dynamic positioning system for model basin facility, and the development of an ice basin numerical simulator. Finally, the opened perspectives are discussed with a special focus on the operational matters.

Abstract The development of icebreaking ships started systematically in late 1960ies. That time also it came quite common to also perform model tests in ice. Model tests are used to make comparisons with alternative designs. Full-scale tests are used for two main purposes; to verify the performance of a vessel built and to create correlation data base to be able better scale up model test results. One of the key issues is the guaranteed performance of a vessel which means that in the shipbuilding contract there is the clause containing for instance the icebreaking capability of the vessel. In uniform level ice it could be a speed and ice thickness. During the last 40 + years the full-scale onboard vessel tests by Aker Arctic and its predecessors; namely W& ärtsil& ä Ice Model Basin (WIMB), W& ärtsil& ä Arctic Research Centre (WARC) and Masa-Yards Arctic Research Centre (MARC), consist of a total of 96 different vessels tested on 140 occasions. These tests include in addition to vessel performance verification also measurements on propulsion loads and hull stress. In addition there have been numerous research projects and projects on vessel operations. This paper will summarize and discuss the extent of work performed in full scale testing concentrating on vessel performance, what and how things have been done.

Abstract Ice-capable jackups can offer mobility and efficiency while drilling in shallow, ice-infested waters. However, the industry has limited experience operating jackups in ice environments, and technology gaps and uncertainties remain when assessing ice loads on multi-legged jackup units. The Graphic Processing Unit (GPU)-based Discrete Element Method (DEM) can be applied to investigate the degree of ice effects on multi-legged jackup units. In previous applications, the DEM method was used to simulate ice interaction with structures, but computations were limited to simple interactions. Introducing a GPU platform enables a substantial increase in the number of elements characterized so that the interaction of multi-legs and ice can be modeled with greater efficiency. For example, level ice modeled with 3-D bonded spherical elements can address the buoyancy, current effects and drag force encountered in harsh environments. Additionally, the parallel bonding approach and de-bonding criteria are used to model level ice freezing and breakage. In this investigation, the global ice loads on a jackup unit's single leg are analyzed to determine the effects of ice thickness, ice strengths and drifting velocities. The estimated ice loads are validated against full-scale data acquired from one application in Bohai Bay. Further, the mechanism of the interaction between the multiple legs and level ice is analyzed using the large-scale GPU-based simulation, the effects of sheltering and jamming are investigated and the ice loads simultaneously acting on the multiple legs are estimated to determine the total shear force and overturning moment.