PO049

The numerical simulation of a floating wind turbine. A comparative study.

Cian Desmond 1, Rachel Chester2, Simon Watson2, Jimmy Murphy1
1University College Cork, Cork, Ireland, 2Loughborough University, Leicestershire, UK

Abstract

Numerical simulations considering a semi-submersible floating wind turbine platform are performed using two industry standard numerical design tools, OrcaFlex and Flexcom. An inter-comparison is performed in order to assess simulation quality with respect to key floating wind turbine design parameters:

• Fairlead mooring line loads.
• Hub height acceleration.
• Platform pitch.
• Platform Response Amplitude Operators.

Validation data from physical tank testing at a scale of 1:36 is used to benchmark the numerical simulations. These tank tests were conducted at a scale of 1:36 in the Deep Ocean Basin of the Lir National Ocean Test Facility, Cork, Ireland.

The floating platform considered in this study is designed to support the NREL 5MW reference turbine as part of the LEANWIND EU FP7 project. The lightweight semi-submersible platform consists of a triangular structure with three vertical rectangular towers connected below the waterline by horizontal pontoons of rectangular cross section. The platform is buoyancy stabilised by its water plane inertia allowing for a relatively shallow draft and high stability by minimising the influence of wave excitation forces. Water ballast within the corner towers and along the submerged base act as a permanent ballast to lower the platform to its operational draft. The platform is moored using a three-point centenary system.

 

Method

OrcaFlex and Flexcom are offshore structural analysis software packages originally developed for the oil and gas industry. Both codes use the finite element method to perform a coupled dynamic analysis of mooring lines connected to a floating body.

Tank testing was conducted using a Froude scaled model of the LEANWIND platform. A Froude scaled catenary mooring lines was used for the primary mooring line in order accurately replicated the system dynamics.

A total of 15 regular and irregular sea states were considered with significant wave heights ranging from 1.5 m to 5.5 m. Static wind loading was applied. These conditions were examined using both numerical tools. Results for key wind turbine design parameters were compared to the analogous tank test data.

 

Results

There was generally a high level of agreement in the inter-comparison between Flexcom and OrcaFlex throughout this study and a reasonable agreement with the validation data.

Some shortcomings were identified in the calculation fairlead loads. This has been attributed to the absence of mid-frequency responses between the second-order hydrodynamics and the wave spectrum in the numerical models during wave loading. This phenomenon was not observed for the simulations undergoing applied wind, which suggests that aerodynamic loading dominates the fairlead loads and overpowers this mid-frequency range in the hydrodynamics.

The numerical models consistently underestimated the acceleration at hub height. Given the importance of these motions in the design of turbines. This is a key consideration for industry.

 

Conclusions

The analysis presented in the full paper will show good agreement between two industry standard design tools for the consideration of floating body dynamics when considering a semi-submersible floating wind turbine platform. Particular instances when results diverge will be highlighted and possible causes will be identified. Analysis will consider key design parameters of direct relevant to the offshore wind energy industry.

The analysis will also underline the importance of physical tank testing in supporting and reducing risk during the iterative design process in conjunction with numerical codes. Limitations and synergies between numerical and physical modelling approaches will be identified.

 

Objectives

Attendees will be shown the relevance of OrcaFlex and Flexcom in the design of floating offshore wind turbines. The industry is increasingly moving towards coupled aerodynamic - hydrodynamic models such as FAST. Whilst the ability of these codes to give a greater appreciation of the system dynamics is indisputable, it come at considerable computation cost and requires a well-trained operative to ensure simulation fidelity. OrcaFlex and Flexcom offer a more attainable solution, however certain limitations must be noted.

The presentation will include video footage of both tank testing and numerical simulation in order to stimulate discussion.