Professional Documents
Culture Documents
Number:
Area:
Authors:
HYDRODYNAMICS
Daniel Cueva / Fernando Faria
Summary
1.
BACKGROUND ........................................................................................................................... 3
2.
3.
2.1.
2.2.
3.2.
3.3.
3.4.
4.
5.
5.2.
6.
7.
ANTI-ROLL TANKS............................................................................................................. 18
7.2.
8.
REFERENCES ............................................................................................................................ 23
9.
1. BACKGROUND
In the last decades, several new technologies were developed in order to fulfill the increasing
demand for deep-water offshore production units. As one of the most successful, the FPSOs
(Floating, Production, Storage and Offloading) were able to mix all the experience in ship operation
with a reliable vessel for oil production.
The storage capacity and the safety of a big water plane area hull changed the scenario in the
offshore industry, and the concept becomes the standard solution for several applications all over
the world. According to a 2007 survey, there are 113 FPSOs currently in operation and 33 in
construction phase (ref. [26]).
However, the direct application of the technology also introduced new problems. Since most of the
current fleet of FPSOs are composed by converted units, several aspects that should be considered
in an offshore design are left behind, for instance motions in waves.
Excessive roll motions and accelerations are one of the key aspects in FPSO analysis, since they are
directly linked with downtime. The excessive motions of a FPSO are caused not only by severe
environmental conditions, but also due to mild combinations, that may lead the unit to unfavorable
beam seas conditions, in both SMS (Spread Mooring System) and SPM (Single Point Mooring).
2. MOTION CAUSES
2.1. ENVIRONMENTAL ASPECTS
The SMS system is defined by the presence of several mooring lines, which does not allow the unit
to weathervane. To weathervane means that the ship can rotate in the horizontal plane (yaw) into
the direction where environmental loading due to wind, waves and current are minimal.
As a result, SMS are usually applied where the weather conditions are moderate and the current
direction is relatively fixed. However, since the heading is almost permanent, the FPSO will not
encounter head waves all times, and deviations from 15 to 30 degrees are expected. Another
important issue is that, in most cases, the main driver to the heading definition is the mooring loads.
It means that the fixed heading of a SMS system will be specified by the optimum arrangement of
the lines, which is not necessarily the optimum for motion characteristics.
Besides that, the swell conditions are also present, and may result in long period waves hitting the
side of the unit.
In the SPM system a turret is usually used, and the system is able to weathervane about the mooring
point. As a consequence, the unit will be able to rotate and the final equilibrium position will result
in a minimum load in the mooring system, when exposed to waves, wind and current.
However, even in SPM system critical sea conditions may appear, resulting in large roll motion
amplitudes. In Brazil, for example, simultaneous environmental conditions result in a difference of
wind and wave incidences up to 90 degrees, such as Wind from East and Swell from South. When
the wind speed is high enough to induce the units heading, beam sea conditions will be faced by the
vessel.
Table 1 shows a summary of the sea states for Campos Basin for which the wind speeds were above
5 m/s in order to simulate the conditions for which the vessels heading starts to be affected by the
wind. Beam sea states are defined as conditions for which the angle between wind and wave is in
the vicinity of 90 degress.
Table 1 - Wave conditions
Return Period
1 month
1 year
10 years
100 years
Comparing the values presented above and the metocean information from Table 1, it is possible to
conclude that resonance will be induced, and high amplitudes will be achieved, especially for ballast
condition. Recent full scale monitoring showed that roll motion in FPSOs may achieve 15 degrees
(single-amplitude) in certain conditions (ref. [11]).
Loading and offloading operations are also critical for roll motions, since the ships cargo sequence
has direct influence on the metacentric height and the roll restoring moment. It is always desirable
to include intermediate loading and offloading situations on the motions analysis, in order to verify
possible problems.
3. MOTION CONSEQUENCES
The presence of excessive roll amplitudes is responsible for several aspects of the FPSO design and
operationability.
3.1. COMFORT ON BOARD
First point that should be evaluated is the crew performance due to comfort when subjected to lowfrequency vibrations, generally imposed by vessel motions. This oscillation may result in motion
sickness, body instability, fatigue, discomfort and increased health risk. A cumulative measure of
exposure to low-frequency oscillation may be used to provide an indication of the probable
incidence of motion sickness. ABS (ref. [1]) defines the vertical Motion Sickness Dose Value MSDVZ,
in m/s1.5, by the following expression:
Where azw(t) is the z-axis acceleration as a function of time in meters-per-second squared (m/s2),
weighted by the Wf frequency weighting as defined in BS 6841:1987 and ISO
8041:1990/Amd.1:1999, and T is the duration of the motion in seconds.
The measurements must be taken in 0.1 to 0.5Hz frequency range (2.0 to 10.0s period), and the
following criteria should apply.
Table 2 - Confort criteria
Table 3 shows the expected accelerations for a process module in a converted FPSOs due to roll
motions in extreme sea states (100 years return period; Campos Basin), based on BV formulations
(ref. [25]).
Table 3 - FPSO topside accelerations
Longitudinal
(m/s2)
(g)
Transversal
(m/s2)
(g)
0.985
4.457
0.100
0.457
Vertical
(m/s2)
(g)
-12.362
-1.260
Some alternatives may be considered in order to allow the use of SCRs in FPSO, such as the
improvement of SCR fatigue and strength performance, the reduction of the FPSO motion or the
decoupling of riser and vessel motion, through the use of submerged intermediate structures (riser
towers, submerged buoys, etc) (ref. [18]).
4. ROLL HYDRODYNAMICS
First order motions of ships can be obtained by a simple dynamic equation that computes values of
inertia, damping, wave and restoring forces.
A typical roll motion with single degree of freedom can be described as (ref. [36]):
M t
(1)
Where is the roll angle, and are the first and second differentiations with respect to time, i.e.,
angular velocity and angular acceleration. is the mass moment of inertia in roll (considering its
additional part), is the nonlinear damping moment, is the nonlinear restoring moment, and
M(t) is the exciting moment by waves.
Various expressions of damping and restoring terms were used to simulate the nonlinear
characteristics of roll motion. The most commonly used representations are:
B
C
BL
C1
BN
C3
(2)
where BL and BN are the linear and nonlinear damping coefficients, respectively, and C 1 and C3 are
the linear and third-order restoring moment coefficients.
The usual procedure for calculation of the ship motions is the use of the so-called diffractionradiation software based on the potential flow theory. The damping calculated by this kind of codes
is linear and related to the wave making.
For linear assumption of the phenomena the expression (2) reduces to:
B
C
BL
C1
(3)
10
5. DAMPING PROBLEM
Linear damping coefficient, or potential damping coefficient, is the one obtained from potential flow
theory and depends on the hull geometry only. Other types of damping that are not evaluated by
this theory must be approximated or obtained from model tests.
These damping sources are identified as a dissipation of energy due to the drag, friction, flow
separation and some other effects.
Its influence is usually obtained from two common procedures (ref. 28):
Free oscillation model tests, or decay tests, provides the system natural period and the damping of
the unit. It is possible to say that these tests are the most reliable tool for predicting roll damping of
a unit. Several cares must be taken in order to correct evaluate damping from the time series
obtained from these model testing, as explained below.
Semi-empirical methods are based on an extensive series of model tests which results are fitted to
an empirical formula obtained after combining the model test results with some theoretical
considerations. The Ikeda-Himeno is the most popular method for evaluation of roll damping for
ship like bodies. This method decomposes the total roll damping in several components that
represent different physical phenomena occurring on the hull. The numerical implementation of this
method can be obtained from [13].
Unfortunately it is known that the roll motion is nonlinear. That means that the methods explained
above do not evaluate precisely the correct damping of the unit and other tests are often done in
order to enclose this wave amplitude influence. Regular wave tests, decay tests in waves or forced
roll motions tests in waves are some of the ones that comprises this behavior.
An example of nonlinear FPSO roll damping, due to different wave amplitudes, is shown below. The
hull was tested under regular and irregular waves, and its results were then compared to a
numerical calculation by a diffraction-radiation software.
11
Ones can see that different external values of damping were included in the numerical calculations
in order to correctly adjust the nonlinear damping that showed up for the several waves tested over
the body.
5.1. FREE OSCILLATION TESTS
The principles of the method are quite well known. An initial displacement in roll is given to the body
and the time history of the motion is registered.
(4)
p1
p2
(5)
p1
By assuming a solution of the form
(6)
st
(7)
12
x x 0 exp
sin
(8)
x k and x k
ln x k
ln x k
(9)
which gives:
(10)
xk
2
ln
Tm
xk
where Tm is twice the period between
p1
x k and x k
16 x k
p
3 Tm 2
1
(11)
Plotting the points obtained and fitting a straight line to them, the values of p1 and p2 are found.
13
It is shown below two examples of simulations where good and bad results were obtained. The cases
were simulated for the same body with different irregular wave conditions. Each condition had its
damping adjusted properly.
As displayed, the first numerical simulation adjusted well its numerical time series to a time series
obtained from the same wave on an ocean basin. The second time series, however, could not
achieve the same results.
14
(12)
in which all involved quantities in the integrand are of the first order as for the free-surface
elevation, for the velocity potential, X T + R r 1 , 2 , 3 for the displacement due to the
translation T 1 , 2 , 3 and rotation R 1 , 2 , 3 , and r x x0 , y y0 , z z0 for the
position vector with respect to the reference point x0 , y0 , z0 of rotation. In (12), stands for the
15
intersection (waterline) of the hull H at its mean position with the mean free surface F z = 0 . The
normal vector n n1 , n2 , n3 is positively oriented inwards to the fluid.
Another formulation based on the momentum theorem has been developed by Maruo (1960) and
extended to the moment around the vertical axis by Newman (1967). This formulation involving first
order wave field in the far field is often called far field formulation and it's preferred in practice as it
provides good convergence and stability.
The momentum formulation, equivalent to one given in Newman (1967), is given as:
(13)
written on a surface S located at infinity and its upside boundary touching vertically the mean
free surface.
Although both formulations presented above should give equivalent results, the near-field
formulation suffers from poor numerical convergence, as the singularities are present in the velocity
field around the hull area with a sharp variation of geometry like corners. On the other hand, the farfield formulation is less sensitive to the hull mesh discretization and is numerically robust. However,
the far-field formulation only gives the three horizontal components of loads (ref. [33]).
Under the practical point of view, we may find roll second-order motion amplitudes higher than
first-order, especially when working in high natural period vessels. It is important to highlight that, in
many cases, all numerical analysis are carried out considering only first-order response and
horizontal second-order loads (for mooring and riser evaluation). As a result, the second-order
motions are only going to be verified during model tests, when the flexibility for major modifications
is smaller.
As an example, Figure 14 presents the experimental spectral roll response for a FPSO design with
high natural period. Ones can see that the units response is located far away from the wave peak
period spectrum, which is attributed to nonlinear mechanisms. However, no linear response (first
order) is observed (ref. [12]).
16
17
7. SUPPRESSION SYSTEMS
The non-dimensional roll RAO for a bare hull can be 10 times greater than the wave slope. For that
reason, a control of this roll motions is sometimes needed or desirable. The almost total lack of
inherent roll damping means that small additions to this damping can produce large reductions in
the response.
Since the most severe roll motions occur at resonance, the best way of reducing it is to increase
damping. The most common means of doing so is the installation of bilge keels or, if more control is
desired, the use of special anti-rolling devices.
18
A RAO of a ship in beam seas with and without a stabilizer defined by the parameters is shown
below.
UNSTABILIZED:
Represents a RAO for a
ship without anti-rolling
device, with a roll damping
ratio of 0.05;
STABILIZED:
For a ship with a good
passive roll tank, for the
parameters shown above;
Figure 16 - Comparison of the computed roll response of a typical ship with and
without a good passive roll tank
Controlled-passive tanks are similar to passive tanks, with the exception that the area of the water
cross-over tube is larger in cross section than in the passive one and that it contains a valve that
controls the air flux through the air cross-over duct. In general, this kind of system controls natural
and low frequencies roll motions. Figure 17 shows the comparison between different types of
passive tanks.
VALVES OPEN:
Represents a RAO for a
ship stabilized with a
controlled-passive tank
that is not operating;
VALVES ACTIVE:
For a ship with a working
controlled-passive tanks;
PASSIVE TANK:
For a ship with a good
passive tank;
19
Active tanks generally imply that the system requires the use of machinery of significant power and
is designed to be much more effective in eliminating roll motions than passive systems. For that
reason it consists on a system that detects motions of the ship and predicts the roll moment that will
be applied on an immediate future. With that information, the system applies a roll moment that
will cancel this predicted moment. For U-tube tanks this kind of system usually provides a large lag in
the tank response and requires a great amount of efficiency and power from the pumps for the
system to work properly.
For FPSOs, however, the biggest problem is related to the draft variation. It makes the application of
any anti-roll stabilizing tank a very difficult task, since the natural period of the U-tank system would
have to be tuned all time, changing it from a passive system to an almost active system, with all its
disadvantages. Additionally, any mistake in the anti-roll tank tuning may lead even to an
amplification of the roll motion (ref. [11]).
7.2. BILGE KEELS
The magnitude of roll depends both on the relationship between the FPSO and wave dimension, as
well as the resonant effects. Since modifications on the geometry are usually complicated, especially
in converted units, the influence of damping becomes fundamental. However, the ship shape and
the length over breadth relation results in a small damping in transverse motions, with consequent
great sensitivity to resonant effects. Thus, the first attempt to reduce roll motion can be the
introduction of artificial damping, and bilge keels are the simplest way to do it.
These appendages are usually constructed from flat plates that form a sharp obstruction to the roll
motion. In ships, the height of a bilge keel is usually selected to be such that the tip of the bilge keel
lies within the maximum beam and above the baseline, in order to keep it protected during docking,
drydocking and shallow water sail. The size of the bilge keel also has to be restricted in ships, since it
can increase the forward resistance. Figure 19 presents a comparison of bilge keel effects with
forward speed.
Formulations for damping estimation, based on the model test regression analysis were developed.
Equation 14 presents the zero-speed damping ratio estimation, based on ship and bilge keel
characteristics.
20
(14)
Where:
ABK
bBK
CB
d
L
B
T
4
Although the bilge keel effectiveness increases as larger the structure is, there is a limitation due to
the increasing in the advance resistance. However, since FPSOs are stationary systems, there is no
limitation to bilge keel dimensions, which could be enlarged, resulting in two different hydrodynamic
effects: the increase of damping forces and the increase of added inertia.
Several studies had been carried out in order to evaluate the effectiveness of large bilge keels for
FPSO, mostly based on model tests, and results have shown that enlarged keels can produce more
than 100% increase on damping coefficients. Table 4 shows the damping, as a percentage of the
critical damping, for different sizes (widths) of bilge keel (ref. [11]).
Table 4 - Enlarged bilge keel effectiveness
Damping
2.46%
3.90%
6.51%
However, studies have shown that changes on the width are much more effective than modification
on the extension, in terms of extra damping. Some investigations have already discussed about
adverse pitch motions when large bilge keels are extended too far forward.
Attention must be paid for possible effects of added inertia when it comes to very large bilge keels,
which can result in higher natural periods. Table 5 shows the roll natural period variation on a newbuilt FPDSO concept after the introduction of a 4m wide bilge keel (ref. [24]).
Table 5 - Natural period variation due to enlarged bilge keel
20.74
23.20
21
As previously highlighted, higher roll natural periods may be useful in terms of detuning the motion
response against the wave energy spectrum, but may lead to undesirable second-order effects.
However, since the motions cased by second-order effects are slow (and drag forces are related to
squared velocity), the effectiveness of bilge keels are smaller.
Additionally, extra-large bilge keels are subjected to high drag forces, which may induce to structural
and construction problems.
22
8. REFERENCES
[1] ABS; Passenger Comfort on Ships, 2001, Houston, USA.
[2] API RP 2SK; Design and Analysis of Station Keeping Systems for Floating Structures; Third
Edition.
[3] Buchner, B.; Green Water on the Bow of FPSOs, Hydrodynamics of Floating Structures Training
Course, 2007.
[4] Bunnik, T.; Cozijn, J.; Analysis of Mooring Systems, Hydrodynamics of Floating Structures
Training Course, 2007.
[5] Chakrabarti, S. K.; Hydrodynamics of Offshore Structures, 1987, Computational Mechanics
Publications.
[6] Chakrabarti, S. K.; Offshore Structure Modeling, 1994, World Scientific Publishing Co.
[7] Chen, X-B.; Orozco, J-M.; Malenic, S.; Evaluation of Wave and Current Loads on Offloading
FPSOs, 2005, OTC 17180, Houston, USA.
[8] Cueva, D.; Campos, F.; Donato, M.; Ferrari, J.; Torres, F.; Nishimoto, K.; Dimensional Study for
Brazilian FPSO, 2005, OMAE2005-67333, Halkidiki, Greece.
[9] Del Vecchio, C.; Costa, L.; Station Keeping in Deep and Ultradeep Waters, 1999, OTC 10778,
Houston, USA.
[10]Faltinsen, O. M.; Sea Loads on Ships and Offshore Structures, 1990, Cambridge University
Press.
[11]Ferrari, J.; Ferreira, M.; Assessment of the Effectiveness of the Bilge Keel as an Anti-Roll Device
in VLCCSized FPSOs, 2002, ISOPE, Kitakyushu, Japan.
[12]Ferreira, M.; Torres, F.; Cueva, D.; Ceppollina, D.; Pinheiro, S.; Correa Jr, H.; Umeda, C.;
Hydrodynamic Aspects of the New Build FPSOBR, 2005, IWAOH, Rio de Janeiro, Brazil.
[13]Ikeda, Y.; Himeno, Y.; Tanaka, N.; A Prediction Method for Ship Roll Damping, 1978, Report of
University of Osaka.
[14]Kaster, F.; Rossi, R.; Masetti, I.; Falkenbers, E.; Karlsen, S.; Waclawek, I.; DICAS A New Mooring
Concept for FPSOs, 1997, OTC 8439, Houston, USA.
[15]Lapidaire, P.; Leeuw, P.; The Effect of Ship Motions on FPSO Topsides Design, 1996, OTC 8079,
Houston, USA.
[16]Lee, C.; On the evaluation of quadratic forces on stationary bodies, 2006, Chestnut Hill, USA.
[17]Lewis, E.; Principles of Naval Architecture, 1989, Second Revision Vol. III, SNAME.
[18]Luo, Y.; Ye, W.; Mooring and Riser Design for GoM FPSOs in 10,000 Ft Water Depth, 2005, OTC
17620, Houston, USA.
[19]Maruo, H.; The drift of a body floating on waves, 1960, Ship Res.
[20]Mastrangelo, C.; One Company's Experience on Ship-Based Production System, 2000, OTC
12053, Houston, USA.
23
[21]Neto, T.; Lima, H.; Conversion of Tankers into FPSOs and FSOs: Practical Design Experiences,
2001, OTC 13209, Houston, USA.
[22]Newman, J.; Second-order diffraction in short waves, 2004, Workshop on Water Waves and
Floating Bodies, Cortona, Italy.
[23]Newman, J.; The drift force and moment on ships in waves, 1967, J. Ship Res.
[24]Nishimoto, K; Videiro, P; Fucatu, C; Matos, V; Cueva, D.; Cueva, M.; A Study of Motion
Minimization Devices of FPDSOs, 2001, OMAE2001/OFT-1131, Rio de Janeiro, Brazil.
[25]NR 497 DTM R00 E; Hull Structure of Production, Storage and Offloading Surface Units, Bureau
Veritas, 2004.
[26]Offshore Magazine; 2007 Worldwide Survey off Floating Production, Storage and Offloading
(FPSO) Units, 2007, Houston, USA.
[27] Olgivie, F.; Second-order hydrodynamic effects on ocean platforms, 1983, Proc. Intl Workshop
on Ship & Platform Motions, Berkley, USA.
[28]Orozco, J.; Raposo, C.; Malenica, S.; A Practical Procedure for the Evaluation of the Roll
Motions of FPSO's Including the Non potential Damping, 2002, OTC 14234, Houston, USA.
[29]Palazzo, F.; Silva, A.; Oliveira, C.; Oliveira, M.; On the Latest Petrobras FPSO Design Procedures:
Hydrodynamic and Mooring Aspects, 2004, OMAE-FPSO04-0079, Houston, USA.
[30]Park, I.; Shin, H.; Chung, H. Beek, J; Development of a Deep Sea FPSO Suitable for the Gulf of
Mexico Area, 2002, OTC 13999, Houston, USA.
[31]Portella, R.; Kameyama, V.; Wibner, C.; Maloney, J.; P43/P48 Global Motion and Stability
Analysis: A Compromise Combination to Define the FPSO Operational Behavior, 2003, OTC
15138, Houston, USA.
[32]Portella, R.; Mendes, B.; DICAS Mooring System: Practical Design Experience to Dismystify the
Concept, 2002, OTC 14309, Houston, USA.
[33]Rezende, F.; Chen, X.; Ferreira, M.; Second Order Roll Motions for FPSO's Operating in Severe
Environmental Conditions, 2007, OTC 18906, Houston, USA.
[34]Santos, A.; Henriques, C.; Pimenta, J.; Improvments Achieved in the Project of FPSO P-50,
2004, OTC 16705, Houston, USA.
[35]Sousa Jr, J.; Fernandes, A.; Masetti, I.; Silva, S.; Kroff, S.; Nonlinear Rolling of an FPSO with
Larger-than-Usual Bilge Keels, 1998, OMAE98-0412, Lisbon, Portugal.
[36]Wu, X.; Tao, L.; Li, Y.; Nonlinear Roll Damping of Ship Motions in Waves. 2005, ASME 205-211.
24
STABILIZER SIZE
GM T
GM T0
t
t
n4
2 g
'
S
n4
'
0
A0
A
Bt
Bt
B ct
25
STABILIZER CAPACITY
S''
'
S
t
L
''
0
q
R
t
t
S
t
0.20
1.08
0.30
0.05
0.00
26