Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Materials: Challenges and potentials' taking place on Thursday, 13 March 2014 at 09:00-10:30. The meet-the-authors will take place in the poster area.

Ngoc-Do Nguyen DNV GL, Germany
Co-authors:
Ngoc-Do Nguyen (1) F P Philipp Gujer (1) Andreas Manjock (1)
(1) DNV GL, Hamburg, Germany

Printer friendly version: printer.gif Print

Abstract

DETERMINISTIC AND STOCHASTIC APPROACH TO CONSIDER DYNAMIC SEA ICE LOADS IN GL/IEC/ISO STANDARDS

Introduction

Many offshore wind farms have been or are being installed in areas where sea ice is of common occurrence, e.g. the Baltic Sea proper. In those areas, dynamic sea ice loads can be a problem for both local and global structures of offshore wind turbines. The calculation of dynamic sea ice loads is difficult. Furthermore, the calculation requirements in standards, as e.g. GL COWT 2012, ISO 19906 and IEC 61400-3 which are based on deterministic ice loads, seem rather conservative and are to be compared with stochastic approach.

Approach

Compared to oil & gas structures, which are rather stiff, offshore wind turbines are very elastic structures. While stiff structures are dominated by extreme sea ice conditions offshore wind turbines are dominated by cyclic ice conditions. Effects like frequency lock-in can become design-drivers for both the substructure and rotor nacelle assembly.
For extreme sea ice conditions extensive literature is available. For elastic structures where cyclic ice loading is relevant only limited information is available. At the time being only few experts have the knowledge of performing proper cyclic sea ice calculations.
Ice data are known from observation, measurements and historical data, collected in ice charts, e.g. in the Sea Ice Atlas. As for wind and metocean data appropriate institutes can deliver such sea ice data, e.g. VTT Technical Research Centre of Finland. Statistical methods are used for long-term data such as extreme ice thicknesses.
Only drift ice has to be considered in the calculations since land-fast ice occurs only in narrow zones along the coast line or far North in Finnish waters. For offshore wind farms in the Baltic proper area first-year level ice is the most probable drift ice type since sea ice does not occur in every winter in these areas. Drift ice has to be considered for both extreme and fatigue loading.
Sometimes ice ridges can occur due to interaction of ice sheets at their edges. Ice ridges have to be considered on addition to drift ice. Ice ridges have to be considered for extreme loading.
In this study, dynamic effects of sea ice on jacket and monopole structures are considered based on the different standards. Due to this effect, fatigue seems to be more critical for monopoles meanwhile extreme loads could occur in both jacket and monopile foundations.


Main body of abstract

Focus of the presented analysis is on dynamic sea ice and in particular on ice induced vibrations. Ice-induced vibrations can arise when the ice sheet acts continuously on a vertical structure. The dynamic ice-structure interaction process is influenced by the ice velocity and the waterline displacement of the structure. Dynamic ice-structure interaction depends, besides the ice speed, on the waterline displacement. For slender, elastic structures as offshore wind turbines the maximum waterline displacement usually corresponds to the second bending eigenmode of the substructure, coupled with the rotor blade eigenmodes. Therefore, ice-induced vibrations can be design-driving for both the substructure and the rotor nacelle assembly, especially the blades!
Intermittent ice crushing, as shown in Figure 1a), can arise if a compliant structure is exposed to ice action at a low speed. The interaction involves a loading phase and an unloading phase. In the loading phase, the structure moves in the same direction as the ice. The ice edge experiences ductile deformations and the ice action gradually increases. The external ice action and the internal forces of the structure are usually in a static balance when the ice action reaches a maximum value. At the peak value of ice action, brittle crushing starts at the ice edge, leading to relaxation vibrations in the structure that decay during the unloading phase. The rate of decay depends on the total damping provided by the soil and the structure (taken from /1/).
Frequency lock-in, as shown in Figure 1b), can occur at intermediate ice speeds, ranging typically from 0.04m/s to 0.1m/s, as the time-varying ice actions adapt to the frequency of the waterline displacements of the structure. The vibrations of the structure are typically sinusoidal in this condition. Similar to intermittent crushing the ice-structure interaction exhibits alternating phases of ductile loading and brittle unloading. The time history of the ice action depends on the characteristics of the ice and the structure (taken from /1/).
Figure 1c) shows typical records of the ice action and the structural response for continuous brittle crushing. This occurs at higher ice speeds, typically when higher than 0.1m/s. Both the ice action and the response are random (taken from /1/).
The described ice break modes are modeled by time-varying external ice force signals and are applied for the load simulations. ISO 19906 gives simplified time histories of the ice force (on vertical structures) due to intermittent crushing and frequency lock-in, respectively, see Figure 1 and /1, A.8.2.6.1.3/ and /1, A.8.2.6.1.3/. These ice force signals can lead to very conservative structural loads.

IEC 61400-3 gives also simplified time histories of the ice force (on vertical and conical structures), both sinusoidal and triangular shaped, see Figure 2 and /2, E.4.6/. IEC makes no distinction between the different ice-structure interaction modes. As for ISO 19906 these ice force signals can lead to very conservative structural loads.

In GL COWT 2012 /3/, the amplitude of the load oscillation may be assumed to be about ¼ of the static horizontal ice load with a mean value equal to ¾ of the sea ice load.
Ice force signals considering the response of the offshore wind turbine are generated e.g. by Dr Kärnä by using a simplified elastic model of the offshore wind turbine and by using the PSII code /4/. The problem with these is ice force signals is the fact that in the load simulation there is no coupling between the ice action and the offshore wind turbines response. It is very difficult to estimate whether the resulting structural loads are conservative or non-conservative.


Conclusion

It is proved in this study that the deterministic approach to consider dynamic effect of sea ice is conservative and there is room for optimisation.
The effects of dynamic interaction of floating sea ice with main deflection modes of monopile and jacket foundations is shown by simulation results taken the full dynamics of an offshore wind turbine into consideration. It appears that the higher eigenmodes of the structure which are characterized by a large modal deflection at sea level (ice impact level) contribute significantly on the overall loading of the entire structure. The eigenfrequencies of those modes often interfere with specific wind turbine component modes such as blades, shaft or yaw system.
The application of external ice force signals in today’s simulation codes, deterministic or stochastic, might include relevant frequencies of the structure in their spectra. However, this approach do not enables a coupled interaction with the elastic structure. The existing simulation codes have to be extended by methods which includes reaction forces of the deflected structure to the floating ice segment in each time step of the simulation. With the theories of the described approaches of the standards the new ice impact would be computed and would lead to a realistic interactive response with the structure. Therefore only codes are capable which perform simulations in time domain in order to describe this high nonlinear behaviour adequately.
Thus, the development of a reliable method for the consideration of dynamic ice loads within the load simulation is important. Such a method would permit the coupling of the structure’s response and the ice action, as it is the case for wind and waves.



Learning objectives
Only few measurements exists which document dynamic interaction of floating sea ice with tall, slender structures such as offshore wind turbines.
Hydrodynamic damping effects involving structural and soil damping are rarely documented. Fully elastic simulations have to be verified in the future by practical experiences in order to achieve reliable codes and standards for calculating dynamic sea ice interaction.



References
/1/ ISO 19906:2011-04
/2/ IEC 61400-3 Edition 1, 2009
/3/ GL-Guideline for the Certification of Offshore Wind Turbines 2012
/4/ Description of the PSSII model, Karna Research and Consulting, Revision C, 2012-05-26