Automatic analysis of floating offshore

structures

David Aller, Alfredo Bermúdez, Marı́a Teresa Cao-Rial, Pedro Fontán, Francisco
Pena, Andrés Prieto, Jerónimo Rodrı́guez, José Francisco Rodrı́guez-Calo

Abstract In the coming years offshore wind energy will be one of the most promis-
ing areas in the renewable power generation field. Achieving the optimum design of
floating platforms requires a rigorous analysis chain to establish the response of the
whole platform under different scenarios. With this aim, we have developed a soft-
ware package that automatically analyzes the feasibility of a floating structure. The
structure of the platform is defined according to a very general set of parameters, al-
lowing us to consider a wide range of designs. The package calls some commercial
applications and some codes developed by us, to complete the analysis process. Re-
turned results include the hydrostatic equilibrium position, hydrodynamic pressure,
RAOs (response-amplitude operators), material costs and static stresses.

1 Introduction

Offshore wind power is one of the most promising fields in renewable energy gen-
eration in the coming years. More than 90% of the worlds offshore wind power is
currently installed in Europe. According to Global Wind Energy Council, offshore
wind represents today about 2% of the global wind power installed capacity, and this
figure will increase to 10% by 2020, with many ongoing projects mainly in Europe,
United States, China and Japan.

David Aller, José Francisco Rodrı́guez-Calo

Centro de Tecnologı́a Repsol, Ctra. Extremadura km 18, 28935 Móstoles, Spain, e-mail: jfro-
[email protected]
Alfredo Bermúdez, Marı́a Teresa Cao-Rial, Pedro Fontán, Francisco Pena, Jerónimo Rodrı́guez
Department of Applied Mathematics. Faculty of Mathematics, Campus Vida s/n, 15782 Santiago
de Compostela, Spain, e-mail: [email protected]
Andrés Prieto
Department of Mathematics. Faculty of Computer Science, Campus Elviña s/n, 15071 A Coruña,
Spain, e-mail: [email protected]

1
2 D. Aller, A. Bermúdez et al.

Offshore wind has a number of advantages compared to on land such as higher

wind speeds and less turbulence, thus generating more energy from fewer turbines,
and usually fewer environmental constraints. Offshore is particularly suitable for
large scale developments near major demand centers represented by large coastal
cities, avoiding the need for long transmission lines to bring the power to these
demand centers, as is the usual case onshore.
The main areas for exploitation are however found far off the coast, in deep
waters, where fixed supporting structures similar to the ones installed on land are no
longer economical. These distant sites mean more difficult sea bottom operations
and higher waves and thus floating platforms are more suitable in these conditions.
Floating platform designs were initially conceived for Oil&Gas industry opera-
tions. Therefore these designs were associated to huge safety factors due to implica-
tions to human safety and to the environment of the failure of such installations. But
floating wind requirements are completely different and thus the major challenge
for offshore wind development today is to continue to bring down costs, developing
designs aimed at minimizing the capital expenditure and operating expenses while
guaranteeing structural integrity and providing suitable operating conditions for the
turbine.
For a formal optimization cycle to achieve the optimal design that minimizes the
cost of produced energy (balancing the produced power and the expenses), it is first
necessary to develop an analysis chain able to:
• Provide the response of the set platform-tower-turbine to different load scenarios
(wind and waves spectra) in a fast and rigorous way.
• Robustly handle changes in design variables.
The response analysis of the whole set to different wave and wind scenarios is usu-
ally quite complex. The basis of floating structures optimization can be consulted
in [3–5, 9]. Unfortunately these tools are not widely available for the companies
working in this field. With this aim, we have developed a software package that
automatically analyzes the feasibility of a floating structure. The structure of the
platform is defined according to a very general set of parameters, allowing us to
consider a wide range of designs. The package calls some commercial applications
and some codes developed by us, to complete the analysis process. The main steps
are:
• Generation of a CAD file of the floating structure.
• Material cost for the whole structure.
• Calculation of the hydrostatic equilibrium position, subject to moorings and wind
force at the top of the tower.
• Calculation of hydrodynamic pressure and RAOs (response-amplitude operators)
considering moorings and wave interaction.
• Structural analysis of the platform, using the previous calculations.
This analysis tool can be used into a multi-objective optimization strategy. This
can help us find optimal designs depending on the placement of future exploitation
fields.
Automatic analysis of floating offshore structures 3

Figure 1 shows a flowchart of the analyzer program. The rest of the paper is
organized as follows: Section 2 describes how to encode geometry and to create a
CAD model of the platform. Section 3 deals with the calculation of the equilibrium
state. The numerical procedures to compute aerodynamic and hydrodynamic loads
are described in Section 4. Section 5 details the structural study.

Fig. 1 Flow chart for the

analizer program: 1) process
geometry data, 2) mesh for the
buoyancy program, 3) wind
force 4) buoyancy program, 5)
geometry in the equilibrium
state, 6) conformal mesh for
structural study, 7) RAOs and
hydrostatic pressure, 8) power
calculation and 9) structural
study.

2 Geometry encoding

In order to consider a geometry encoding flexible enough to model a wide variety

platform designs, we assume that platforms are mainly composed of empty bodies
made of metal sheets, that is, their internal structures are neglected. We distinguish
between three type of objects: pillars, connectors and towers.
• Pillars are bodies of cylindrical section that give the platform the ability to float.
They can have a rectangular or elliptical base and their dimensions and position
in the space are parametrized, as well as their lateral profile, thickness and anchor
points. If they contain water acting as ballast, water height is also a parameter.
• Connectors have also cylindrical section; their geometry is parametrized in the
same way than pillars. They can connect pillars or other connectors; contact
points with the connected objects are also parameters.
• Towers are cylindrical objects on the top of some pillars; they are intended to
hold wind generators.
The previous information is stored in a file keeping the structure of the three ob-
ject types. Thus, it is very natural not only to change a specific parameter in the file,
but also to remove or include a complete pillar, connector or tower. Such actions are
4 D. Aller, A. Bermúdez et al.

among the first rules than an optimization algorithm based on grammatical evolution
could need to implement (see [10]).
The first step in the analyzer program is to create a CAD model of the platform
from the geometry encoding. To this end, a Python script was programmed to take
advantage of the Python scripting for Rhinoceros [2]. The resulting geometry is
composed of NURBS surfaces that can be exported in several formats. Figure 2
shows the resulting CAD file for the a semisubmersible platform designed by Mitsui
Engineering & Shipbuilding Co. [6].

Fig. 2 Example of a semisub-

mersible platform designed
by Mitsui Engineering &
Shipbuilding Co. Some of
the geometry parameters are
detailed: pillars in blue, con-
nectors in red and towers in
green.

3 Buoyancy position

The hydrodynamic behavior of the platform is modelled with WAMIT [8], which
assumes that the structure is given in the equilibrium state. We have implemented
the calculation of such equilibrium state for a rigid body subjected to its weight,
buoyancy forces, moorings, wind forces applied at the top of the tower and ballasts.
We remark that the movement of a rigid body can be decomposed into the movement
of the center of mass and the movement induced by the rotation respect to the center
of mass. Besides, the total force applied to the body produces a change in the linear
moment, whereas the total moment respect to the center of mass changes the angular
moment of the body. When the body is balanced, both linear and angular moments
are null as well as both the sum of forces and the sum of moments.
To find out the equilibrium state requires to solve a nonlinear system: the condi-
tion of the vertical alignment of the center of mass and the buoyancy center gives two
equations; the balance between total forces and weight gives another one. Among
all possible solutions, only those which are stable are relevant. A position is stable
Automatic analysis of floating offshore structures 5

when the body recovers its position subjected to small perturbations. To calculate
the stable positions, the time-dependent dynamic problem is solved, integrating the
equations of the rigid body with frictional force in a time interval long enough.
We assume that moorings are composed of chains or cables that partially lay on
the seabed. They are modeled with a nonlinear uni-element model based on catenary
(see [11]). Both flexural rigidity and friction with seabed are neglected.

4 Hydrodynamic and aerodynamic modelling

The hydrodynamic interaction between surface waves and the platform has been
computed using the software package WAMIT [8]. Its implemented model is based
on a linear model where a potential representation is applied to the fluid velocity
field. Once this potential is split taking into account the radiation and diffraction
contributions, the hydrodynamic loads on the wetted body surface are computed.
The numerical solution involves the discretization of an integral equation whose
Green function satisfies the free-surface boundary condition. The high-order im-
plementation of this numerical method, the so-called panel method, represents the
surface body geometry by means of continuous B-splines. This geometric setting
is accomplished since NURBS surfaces are approximated by B-splines when the
original structure representation is exported from Rhinoceros.
These numerical simulations allow to evaluate physical quantities such as the to-
tal force and total moment acting on the rigid solid and also fluid fields (pressure,
velocity, and free-surface elevation). However, only the RAOs and the hydrody-
namic pressure computed by WAMIT are relevant for our analyzer. The six RAOs
are transfer functions associated to each degree of freedom (DOF) of the platform
motion. They depend on the heading angle and the frequency of the incident plane-
wave excitations.
For the aerodynamic modelling, the software package FAST [7] has been used
to compute forces and moments induced by wind at the top of the tower. Since
only static wind loads have been considered, only the module Aerodyn was used.
This computational code requires two kind of input data: those ones related to the
platform dynamics (such as the turbine configuration, its weight, characteristics of
its mechanical components, the tower dimensions, its vibration modes, etc.), and
those data related to the aerodynamic setting, which include the physical parameters
of air, wind speed and direction, airfoil profiles and blade configuration. This code
has been used twice in the analysis process (see Fig. 1): Firstly, forces and moments
are computed at the equilibrium position of the platform, which have been taken into
account to determine the buoyancy position; Second, for each frequency considered,
the aerodynamic forces and moments are computed and used as input data in the
structural analysis performed by Code Aster [1].
6 D. Aller, A. Bermúdez et al.

5 Structural study

The structural analysis is done using Code Aster [1], a finite element code which
includes a wide variety of mathematical models. Pillars are modeled as shells, while
connectors and towers are modeled as beams with a shell transition at the end. Since
Code Aster can be executed through Python scripts, it is suitable to be integrated in
the analyzer.
Mesh produced by Rhinoceros is not a conformal one; we created an appropriate
mesh from the CAD file using SALOME. The forces and loads calculated through
the process are provided to Code Aster to define two problems:
• The dynamic problem considers time dependent forces and it is solved in the
frequency domain. Its solution is added to the solution of the next problem.
• The static problem takes into account the hydrostatic pressure, moorings and the
wind force. This is a pure Neumann problem because static forces are balanced
and there are no fixed nodes. In order to remove rigid movements and to have
a well posed problem, assembly matrices are modified and transferred to the
external solver UMFPACK. The result is injected again in Code Aster.
The resulting stress is processed to detect critical values in the structure (see Fig. 3).

Fig. 3 Von Mises norm of

the stress calculated with
Code Aster.

6 Conclusions

• The increasing development of offshore wind power requires tools to evaluate

the validity of the platforms.
Automatic analysis of floating offshore structures 7

• An analyzer program is presented in this paper, being the result of a project con-
ducted by the Repsol Technology Center.
• This program combines commercial and specifically developed software to cal-
culate power, RAOs and stress for each structure.
• Inputs and outputs have been designed to easily integrate analyzer in an opti-
mization code.

Acknowledgements Authors would like to thank the support of Repsol in this project and, in
particular, the enthusiasm showed by Rosana Plaza Baonza and Jesús Garcı́a San Luis. We want
to recognize the work of Ibán Constenla Rozados in the early part of the project. Finally, we
appreciate the opportunity given by the math-in network of presenting this paper in ECMI 2014.

References

1. Code Aster: Analysis of Structures and Thermomechanics for Studies & Research. URL
http://www.code-aster.org
2. Rhinoceros: Modeling tools for designers. URL http://www.rhino3d.com/
3. Aubault, A., Cermelli, C., Roddier, D.: Parametric optimization of a semi-submergible plat-
form with heave plates. In: Proceedings of the 26th International Conference on Offshore
Mechanics and Arctic Engineering: OMAE2007 (2007)
4. Birk, L., Clauss, G.F.: Automated hull optimisation of offshore structures based on rational
seakeeping criteria. In: Proceedings of the Eleventh (2001) International Offshore and Polar
Engineering Conference Stavanger, Norway, pp. 382–389 (2001)
5. Birk, L., Clauss, G.F.: Optimization of offshore structures based on linear analysis of wave-
body interaction. In: ASME Conference Proceedings. Estoril, Portugal: ASME, pp. 275–289
(2008)
6. Bossler, A.: Japan’s floating offshore wind projects: An overview. Maine Ocean & Wind
Industry Initiative: MOWII Webinar (May 2013)
7. Jonkman, J.: FAST - NWTC Computer-Aided Engineering Tools. URL http://wind.nrel.gov/
designcodes/simulators/fast/. Last modified 28-October-2013; accessed 30-March-2014
8. Lee, C.H., Newman, N.: WAMIT - Wave Analysis at MIT. User manual version 7.0 (2013).
URL http://www.wamit.com
9. Lee, J.Y., Clauss, G.F.: Automated development of floating offshore structures in deepwater
with verified global performances by coupled analysis (2007)
10. O’Neill, M., McDermott, J., Swafford, J.M., Byrne, J., Hemberg, E., Brabazon, A., Shotton,
E., McNally, C., Hemberg, M.: Evolutionary design using grammatical evolution and shape
grammars: Designing a shelter. International Journal of Design Engineering 3(1), 4–24 (2010)
11. Sheng, W., Lewis, T., Alcorn, R.: Numerical investigation into hydrodynamics of moored
floating wave energy converters. In: Proceedings of the 9th European Wave and Tidal Energy
Conference (EWTEC) (2011)