Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
4th M. A. Alyacouby
dept. name of organization (of Aff.)
name of organization (of Aff.)
City, Country
email address
Abstract—This paper presents an alternative approach to in solving the sea keeping problem of FPSO mooring system
predict static and dynamic responses of FPSO-Tanker with during production operation. The vessels are connected to each
offloading model using Neural Network Model. The integrated other by a polyester cable called hawser during the offloading
Neural Network workflow was applied to a complex dynamic
interaction of FPSO for the prediction of its responses during operation which limits the relative movement of the vessel.
offloading. A fit-for-purpose program was developed to evaluate Design of an offloading procedure requires determination of a
the three-dimensional static equilibrium of floating structure system that is safe as well as stable which does not experience
such as FPSO, considering the attached mooring systems. To large motions. The design variables that are pertinent to the
demonstrate the effectiveness and robustness of an Artificial offloading procedure are relative vessel headings, to minimize
Neural Network procedure developed for the prediction of the
responses of moored FPSO. Floating Structure analysis divide the effects of yaw instability, should fishtailing occur, hawser
into two parts, first is to calculate the hydrodynamic coefficient, tension, to avoid failure of the connections, hawser azimuth,
and the second part, is calculate the response in time domain or direction to avoid collision during periods of transient
history. Database is used to record the hydrodynamic coefficients weather and as the relative draft of the vessels increase and
such as damping, added mass, as well as other parameters such decrease distance between the vessels, to avoid collision, and
as mooring stiffness, mooring configurations, mooring angles,
mooring types, hawsers stiffness, viscous damping coefficient, tug or thruster force, as the net magnitude and direction of
etc. A Neural Network Method has capability to predict the the stabilizing forces affect the other design variables and
solution in time domain problems and can be combined with ultimately determine if maintaining equilibrium of the system
Genetic Algorithm to do the optimization. When using artificial is feasible with the use of tugs. Forces on the FPSO include the
neural networks methods for the analysis of floating structures, environmental loads, hawser force, and the mooring reactions.
the computational effort associated with time-domain response
simulation may be reduced drastically compared to classic Forces on the tanker include the environmental loads, hawser
solution strategies. Study has been conducted and result show force, and tug force. Hence the effect of these parameters on
that ANN method has a good accuracy. the dynamic response and hydrodynamic interactions of the
Index Terms—Artificial Neural Network; Genetic Algorithm; FPSO-mooring system is intended to be studied by the use of
FPSO-Tanker; dynamic responses; offshore structure; stability simply mooring configuration to incorporate high nonlinearity
of the complex system. The design of each phase needs to be
I. I NTRODUCTION done carefully and well planned as the safety of structure is
A. Background of Study very important in this case. If decisions are taken based on
The deep-sea oil exploration is increasing as the demand effects in only one part of the life cycle, it may do more harm
for the fuel is increasing day by day. Floating Production than good. Hence a thorough life cycle assessment of the FPSO
Storage and Offloading (FPSO) facilities are found to be very - mooring system is required to have a safe and successful
effective for production, storage and offloading of oil and gas oil production and transportation. The methodology consists
in such higher depth oil fields. There is a significant demand of Simulation of dynamic responses of FPSO- Shuttle tanker
system and mooring line tension using physical model testing,
Identify applicable funding agency here. If none, delete this. numerical and software simulations, Validation of numerical
and software simulation, analysis of FPSO- Shuttle tanker exposed to an environmental load without relying on
system. The expected outcome includes Dynamic response and the computationally expensive time domain analysis,
hydrodynamic interactions of FPSO- shuttle tanker system for especially for screening and design purpose.
the Malaysian and Australian deep waters, Effect of truncated 3) Obtaining an acceptable model to predict the stability
mooring lines on dynamic response of FPSO - Shuttle tanker and the safety of this system with respective to time
system. This will bring in huge benefits to UTP as consultant and complexity of the analysis.
for most of the FPSOs in the Region. In this paper presents an
alternative approach to predict static and dynamic responses C. Objective
of FPSO-Tanker with offloading model using Neural Network The main objective of the research is to develop an in-
Model. The integrated Neural Network workflow was applied tegrated simulation model of side by side FPSO – Tanker
to a complex dynamic interaction of FPSO-Tanker for the system that comprises the safety of offloading condition be-
prediction of its responses in specific area, Malaysia, South tween FPSO and Tanker with minimize the collision during
East Asia. offloading. This objective is to be achieved by the deliverable
simulation aspects as follows:
B. Problem Statement
1) To develop a numerical framework for prediction of the
The FPSO - Shuttle tanker system is connected to each
responses of FPSO during offloading system to Shuttle
other by a polyester cable called hawser during the offloading
Tanker.
operation which limits the relative movement of the vessel.
2) To conduct parametric studies and develop optimization
There are two configurations for the offloading system namely
tool to predict the configuration of mooring lines and
tandem offloading and side by side offloading. Side-by-side
hawsers of FPSO - Shuttle tanker system during offload-
offloading is preferable if the relative motions of the vessels
ing.
are properly accounted for. Design of an offloading procedure
3) To validate the above simulation results using commer-
requires determination of a system that is safe as well as stable
cial software, data from literature review and model test.
which does not experience large motions. Safety requires
sufficient clearance between the vessels and other necessary D. Scope of Research
elements such as mooring lines and risers. Safety also requires
The scope of research is described in particular based on
structural integrity of the hawser line and the connections of
the limitations or applicability of several engineering aspects.
the hawser to the vessel bulkheads. The equilibrium of the
The main portions of the current research are laboratory work
FPSO - tanker system determines the safety and feasibility of
and numerical simulation. Therefore, research is limited to the
the offloading procedure. Design of each phase in the FPSO
following scopes;
- Shuttle tanker system needs to be done carefully and well
planned as the safety of structure is very important in this 1) The numerical simulation is performed by using com-
case. Due to size limitations of wave basins worldwide it is mercial computational offshore dynamics software of
necessary to employ statically equivalent truncated mooring ANSYS Aqwa v.18.
and riser systems to test floating systems to be deployed in 2) Laboratory experiment is conducted by using the off-
deep and ultra-deep waters. The equivalent truncated mooring shore lab of Universiti Teknologi Petronas which has
system is designed iteratively, with a trial and error approach. dimension of 12 m length, 10 width, and height of 1.30
Obtaining an acceptable design for the mooring system using m.
a manual iterative approach is challenging and could take from 3) The environment is limited to unidirectional waves and
some days to some weeks of work. The main problem with steady currents in the experimental modeling.
predicting the stability of side-by-side offloading of FPSO- 4) The contributions of risers are not considered in the
Shuttle Tanker system. The mooring line distribution system numerical or experimental modeling.
give a lot of contribution to stability and response of FPSO- 5) Station keeping systems are limited to spread mooring
Tanker system. Obtaining an acceptable method to predict the lines.
stability of this system with respective to time and complexity 6) For the model tests, linear springs are used to represent
of the analysis. The most important factors on FPSO-Tanker the restoring force for the prototype mooring system.
offloading system is the determination of maximum drift or 7) To investigate the effect of mooring line configuration,
relative motion of both vessels as important as forces act on and the effect of the hawsers line system on the FPSO-
the mooring lines. Related to this problem, three issues that Tanker.
need to be resolved are as follows;
II. LITERATURE REVIEW
1) The prediction of responses of FPSO-Tanker System
with side by side offloading system is required to A. Overview
determine the allowable clearance between two vessels This chapter describes the review of the literature and past
including offloading load. work related to the research on responses of floating structure
2) Find a method that can be used to optimize the response either experiment or modeling. The first review is on the
of the offshore structure such as FPSO-Tanker system literatures of the fundamental theory of offshore structure. The
main topic of offshore structure on this thesis is floating struc- mooring lines. Large elongation and nonlinear stress- strain
ture responses as important as its parameters such as natural relationships as typically observed in polyester mooring lines
frequencies, added mass, added damping coefficient, and en- are taken in to account while developing the nonlinear rod
vironmental loads. Furthermore, literatures that provide latest theory model. The results from the nonlinear rod theory were
journals regarding Artificial Intelligent method to predicting systematically compared with original rod theory Tahar et al
and optimizing the offshore structure such as mooring lines, [7]. Compared Neumann-Kelvin and Double Body lineariza-
risers and floating responses. In term of modeling, several tion, the two classes of linearization theories used in ship
previous laboratory experiments are reviewed to understand motion problem to find hydrodynamic coefficients, response
the method used in the experiment as well as procedures and and load. His studies concluded that the Neumann- Kelvin
results. Furthermore, the use of numerical method in the past computation better represented the ship motion problem when
research is evaluated regarding the modeling preferences and the Froude number is high and for slender bodies while Double
the run of simulation. Literatures on the basic calculation Body Linearization technique is better for low Froude number
of floating structure responses are assessed as preliminary and wide displacement ships Kim et al [8]. Calculated the
review prior to the observation of the literatures, which provide added resistance of ship using direct integration method (near
extensively information related to experimental study on the field approach) and momentum conservation method (far field
floating structure coupled motion with mooring lines. Finally, approach) and concluded that depending on ship speed and
the Artificial Intelligence methods as well as predicting and hull form an appropriate method has to be applied Kim et al
optimizing process in the obtaining forces and responses [9].
of floating structure and its interaction between vessels are
presented. D. Multi Body Interaction
The hydrodynamic interactions and mechanical coupling
B. Dynamic Responses of FPSO effects of two large vessels operating side by side, connected
A comprehensive study for numerical simulations of a turret by elastic lines were investigated by Koo et al [10] using a
moored FPSO in irregular waves with winds and currents. He time domain dynamic analysis program and model testing.
derived the equation of motions of such model in the time Three different methods were used for the analysis: Combined
domain using an uncoupled method and solved rigid body full matrix method (CMM), Iterative separated matrix method
and mooring line dynamics separately Wichers [1]. Compared (SMM) and Non- Hydrodynamic interaction method (NHI).
the coupled and uncoupled analysis for a moored FPSO in They concluded that there may be large discrepancy between
harsh environments and suggested that the uncoupled analysis CMM and other two methods. SMM is better than NHI but
results are efficient to be used in the early design phase of the may still significantly over predict or under predict the actual
mooring system. There was relatively good agreement between values. The cross-coupling terms in CMM has significant
the uncoupled and coupled analysis values even though the effect and that is the best method to obtain an accurate result.
maximum values were different Heurtier et al [2]. Developed The motion response of adjacent multiple floating bodies
a reliable method for predicting the extreme response for the was studied using Rankine Panel Method by Kim et al [11].
FPSO using both short term and long-term approach. This The time domain approach used allowed the coupling with
method can be applied to do the structural analysis of both nonlinear external loads. They showed that, in the multiple
newly built and converted FPSOs Zhao et al [3]. developed a body problem, the 3-D effect is important in predicting the
program in time domain for simulating the global motion of frequencies of the resonant free - surface flows between multi-
a turret moored FPSO. They also conducted physical model ple bodies. Fully nonlinear hydrodynamic interaction between
testing to study the vessel motion and mooring tension for non- two 3D floating structures in close proximity is studied using
parallel wind, wave, current and 100-year hurricane condition the quasi arbitrary Lagrangian-Eulerian finite element method,
in Gulf of Mexico Kim et al [4]. Problems encountered based on the fully nonlinear potential theory by Yan et al [12].
during FPSO tandem offloading operations were investigated The results demonstrate that the 2nd order force components
by Wang et al [5]. Time domain simulation and analysis maybe at a similar level as the 1st order component at the
of offloading system was performed using the multi-body frequency near resonant frequency and under such conditions,
mooring software ARIANE 7.0. Hydrodynamic interactions fully nonlinear model must be employed to estimate the forces
between the vessels were studied. Catipovic et al [6] has done under conditions near resonance. Lee at al studied the dynamic
time domain analysis of FPSO using improved stiffness model behavior of FPSO - shuttle tanker system in current, wind and
for mooring line. Improved model is used to evaluate dynamics waves [13]. The static and dynamic stability of the FPSO-
of FPSO moored using polyester mooring line with linearized shuttle tanker system is analyzed based on using the linearized
axial stiffness. Hydrodynamic loads and interactions of FPSO equations of motion in surge, sway and yaw modes. It is
were calculated using HYDROSTAR (2011). found that the FPSO - shuttle tanker system shows a quite
different stability from those of the FPSO and the shuttle
C. Ship Motion Problem tanker. Vessel motions in waves usually considered as a linier
Developed a theory and numerical tool for coupled dynamic system that response linearly to wave excitation. In the time
analysis of a deep-water floating platform with polyester domain calculation, impulsive response function method based
on frequency-domain results is used in dealing with radiation The hydrodynamic coefficients as added mass µij , radiation
forces, and solution of vessel motions in the frequency domain damping λij and wave forces in the frequency domain are
is required. The linear incident wave potential can be written computed using panel program HydroStar developed by Bu-
in a general form as: reau Veritas (2010). Wave diffraction forces and radiation
ζa g kz forces in the time domain can be obtained using IRF method.
φI (x, y, z, t) = e sin (k(x cos β + y sin β) − ωt − σ) According to Newton’s law, motion equation can be expressed
ω
(1) as:
Where where ζa , ω, β and, σ are the wave amplitude, fre-
ξ¨i
F
Mij + µF F
µF L
F
quency, wave number, heading angle and phase shift, respec- ij (∞) ij (∞)
=
tively. The potential in the flow field can be regarded as the FL
µij (∞ Mij + µij (∞) ξ¨iL
L LL