Newapps Tannuri

Return to Session Directory
Authors Name
Name of the Paper
Session
DYNAMIC POSITIONING CONFERENCE

October 2011
NEW APPLICATIONS SESSION
Utilization of numerical simulation tools for aiding

decisions about DP operations
By Eduardo A. Tannuri1, Carlo F.P.S. Campos2, Allan C. de
Oliveira2, Diego C. Corra2, Joo Luis B. da Silva2
(1) University of So Paulo; (2) Petrleo Brasileiro S/A - Petrobras;
Tannuri, E.A., et al.
New Applications Session
Utilization of numerical simulation...
ABSTRACT
This paper focus on the real applications of numerical simulation tools for aiding and
supporting decisions concerning offshore DP operations.
The time domain simulator TPN (Numerical Offshore Tank) contains mathematical
models for multi-vessel dynamics, mooring and risers lines, DP system, propellers and
environmental forces. It has been developed in a long-term research project with
cooperation between the oil state company Petrobras and University of So Paulo
(USP). Several validations of the numerical code have been done by means of model
scale experiments and real scale measurements.
The first case is related to a recently DP converted crane-barge, that also operates for
pipe-laying. Two complex operations of offshore crane installation of large equipments in
moored or fixed platforms have been extensively studied in the numerical simulator.
Different environmental conditions, failures and relative positions between the barge and
the platform were considered. Operational parameters were evaluated, including crane
lines tension, DP power consumption, oscillation of the equipments, and relative
motions. The results were used for defining the maximum wave condition for a safe
operation, the best relative positioning and the required auxiliary cables for reducing the
motion of the equipments.
The second case is related to a DP drilling vessel operating under failure conditions. A
real DP failure situation was fully reproduced in the numerical simulator, and similar
results observed in the real-scale monitoring could be obtained in the simulator. That
incident required the disconnection of the drilling riser, because the heading (among
some others causes) was limited by the angle of the auxiliary lines (kill and choke) of the
drilling system. After that, the simulator is able to be used for evaluating a more
adequate angle for the BOP, considering the typical environmental conditions of the
Brazilian offshore oil fields and common variations of wind conditions in those fields.
Finally, the simulator was used as an important tool for defining new DP lay-outs for oil
shuttle tankers concerning dynamic behavior, holding capacity after thruster failure and
downtime for offloading operations. For the downtime analysis, a procedure for defining
a comprehensive set of environmental conditions was established. Some time-domain
simulations with automated post-processing and a complete set of static calculation
were then used for defining the allowable conditions and the downtime. The downtime
could be used as an important design criteria for DP layout, optimal FPSO heading and
definition of operational safety zone.
MTS Dynamic Positioning Conference
October 11-12, 2011
Page 2
1. Time Domains Simulation Code

The Numerical Offshore Tank (TPN-USP) is a time domain numerical procedure designed for the
analysis of moored and DP offshore systems. The inputs of the simulator are:
Floating body main parameters (dimensions, mass matrix, etc.);

Aerodynamic drag coefficients (following standard given by OCIMF, 1994);
Current coefficients (following standard given by OCIMF, 1994) or hydrodynamic
derivatives;
Hydrodynamic coefficients (potential damping, added mass, first and second order wave
force coefficients) (Pinkster, 1988, Aranha 1994, Wamit, 2000);
Environmental conditions (wave and wind spectra, current);
Mooring and risers system characteristics;
Thrusters characteristics and layout;
DP modes and parameters (Tannuri; Morishita 2006).
The non-linear time-domain simulation runs in a parallel processing computer cluster and outputs
time series describing the motions of the floating unities in six degrees of freedom (6dof),
tensions on the mooring lines and hawser, propellers thrust and power, etc., and a corresponding
statistical summary. 3D-stereo visualization outputs are also available. Figure 1 shows a block
diagram of the simulator.
4th Order Runge Kutta Integration
X, Y, Z, XX, YY, ZZ
Vessel / Body 1
Forces
Current
Wave
Wind
Hydrostatic
Hydrodynamic
DP / Thruster
Mooring
Risers
Connection Lines
Fenders
Visualization
System
X, Y, Z, XX, YY, ZZ
Vessel / Body 2
X, Y, Z, XX, YY, ZZ
Vessel / Body N
Figure 1 TPN block diagram

1.1 Ship Dynamics and Environmental Action Modeling
In this section, for the purpose of illustrating the mathematical model implemented in the TPN,
only the horizontal motions of the ship is considered. The simulator also contains the models for
all 6-dof of the vessel. The low-frequency motion of the system can be conveniently expressed in
two orthogonal reference systems, one being the Earth-fixed (OXYZ) and the other (Gx1x2x3) a
non-inertial reference frame fixed on the vessel (see Figure 2). The origin of this reference frame
is the intersection of the midship section with the longitudinal plane of symmetry of the ship, and it
is taken as the center of gravity (G). Additionally, the axes of reference frame are taken to be
coincident with the principal axes of inertia of the vessel. Bearing in mind these assumptions, the
resulting equations of low-frequency motions are:
M M x M M x x M x C x
M M x M x M M x x C x
I M x M x M x x C x F
11 1 L
F1EH F1T ;
22 2 L
F2 EH F2T ;
11
1L
22
22
2L
26 6 L
66
6L
26 2 L
2L
6L
26 6 L
11
26 1 L
6L
1L
6L
66 6 L
6 EH
(1)
F6T .
October 11-12, 2011
Page 3
where Iz is the moment of inertia about the vertical axis; M is the mass of the vessel, Cij are
damping coefficients, Mij are added mass terms, F1EH, F2EH, F6EH are surge, sway and yaw
environmental (current, wind and waves) and hawser forces, F1T , F2T , F6T are forces and moment
delivered by the propulsion system. The variables x1L and x2 L are surge and sway low-frequency
velocities and x 6 L is the yaw rate.
Figure 2 Coordinate systems

The position and heading of the vessel related to the Earth-fixed coordinate system are obtained
from the following equation:
X L
x1L
cos( L ) sin( L ) 0

YL T( ) x 2 L , with T( L ) sin( L ) cos( L ) 0

x
0
L
0
1
6L

(2)
Wave-frequency (first-order) motions ( X H , YH , H ) are evaluated by means of transfer functions

related to the wave height, namely, the so-called Response Amplitude Operators (RAOs), and
are obtained via numerical methods modeling the potential flow around the hull. Such an
approach is grounded on the assumption of linear response of wave-frequency motions and on
the uncoupling between wave-frequency and low-frequency motions. Therefore, the real position
of the vessel ( X ,Y , ) can be given by:
X X L X H

Y YL YH

L H
(3)
Current forces and moments may be evaluated through four distinct models: (i) OCIMF Model
(OCIMF, 1994); (ii) a cross-flow model; (iii) a hydrodynamic derivatives maneuvering model; (iv)
an Heuristic Model, (Simos et al., 2001). All models include the possibility of considering constant
or time varying current profiles. In the present case the model (i) is considered.
The simulator covers both cases: constant or gust wind. In the latter case, wind spectra are the
usual Harris, Wills and API types. In the present study, a Harris wind spectrum is used.
Wave action can be regular and irregular. For irregular waves the following common
unidirectional spectra are implemented: Pierson-Moskowitz, JONSWAP and Gaussian. First and
second-order wave effects are modeled (see Pinkster, 1988 and Faltinsen, 1990) and wave-drift
damping effects are included, according to Aranha (1994). Both, first and second order
hydrodynamic coefficients in waves are determined by running WAMIT code (WAMIT, 2000).
1.2 DP Algorithms and Propeller Model
Usually, three main classes of algorithms are used in commercial DP systems. A low-pass filter,
called wave-filter, separates wave-frequency components from the measured signals. Such
decomposition is mandatory, as the DP system is able to control only low-frequency motions.
October 11-12, 2011
Page 4
High-frequency motion control would require an enormous power and could cause extra tear and
wear in propellers. Normally, an Extended Kalman Filter is used to perform such a task.
Furthermore, a thrust allocation logic (TAL) has to be used to distribute control forces to the
thruster system. This algorithm guarantees minimum power consumption, generating the required
total forces and moment to keep the vessel position.
At last, a control algorithm uses filtered motion measurements to calculate such required thrusts.
Commercial DP Systems use a Proportional-Derivative (PD) controller, coupled to a wind feedforward term. Such a term enables one to estimate the wind load action on the vessel (based on
wind sensor measurements), thus compensating it by means of the action of propellers.
The simulator includes models for controllable pitch propellers (cpp) and for fixed pitch propellers
(fpp). The model takes into account their characteristic curves, and is able to estimate real power
consumption and delivered thrust. For fpp propellers, dynamics of rotating parts are also
simulated, accounting for the delay between the control command and the propeller response
due to the inertia of the system. Furthermore, for cpp propellers, a maximum pitch variation rate is
defined in order to simulate the governing mechanism that is responsible for the pitch variation. A
detailed description of DP algorithms included in the numerical simulator can be found in Tannuri
and Morishita (2006).
2. Case Study 1: DP Crane Barge

DP crane vessel operations are critical, since the vessel is kept in close proximity with other unit
and large loads are transported in a pendulum configuration. A precise positioning of the cranevessel is required, in order to avoid unsafe relative motions, as well as keep the load being
transported on a stable position. Figure 3 show two typical operations in a FPSO and a jack-up
platform.
Figure 3 DP crane barge operation (left) FPSO ; (right) Jack-up platform

Petrobras has just finished the conversion of a DP crane barge with 6 azimuth propellers, and
several operations are scheduled for the next months. A research program with the University of
So Paulo has been established, in order to evaluate all operations of the barge by means of
numerical simulations. The outcomes of the analysis are the environmental window and the
relative positioning (barge-platform) for a safe operation in each case.
A complete model of the multi-body system is implemented in the time domain simulator (DP
crane vessel, load being transported, mooring and assistance lines, platform), and a large
number of conditions and positioning are verified, previously to the operation itself. Small scale
experiments are used for providing the coefficients for the numerical model and for performing
some validation of the results. Real-scale monitoring campaign will also be done for checking the
predictions. Figure 4 shows the flowchart of the whole project. A detailed discussion about this
project is given in Rampazzo et al., 2011, and a summary will be presented in this section
October 11-12, 2011
Page 5
Figure 4 Flowchart of the project for evaluating DP Crane Barge operation

2.1 Calibration Model-Scale Tests
Seakeeping tests were performed in IPT-SP towing tank using a 1:48 barge scale model, in order
to obtain the Response Amplitude Operator of the barge. Figure 5 shows some test setup
pictures. Finally, a numerical model was provided using WAMIT to obtain the hydrodynamics
coefficients which will be used in TPN. This model was calibrated using the scale model results.
Figure 5 also shows the numerical Heave RAO, after the correction performed with the viscous
o
o
damping obtained on scale tests, for the incidence of 90 and 180 .
Heave Respoonse Amplitude Operator
1.4
90 degrees
180 degrees
1.2
RAO3 (m/m)
0.8
0.6
0.4
0.2
10
15
T(s)
20
25
30
Figure 5 (left) Sea-keeping test setup, (right) Heave RAO calibrated with the tests
Also in the towing tank of IPT-SP, the current coefficients were obtained by means of a captive
test with the hull. All heading angles were considered (from 0 to 350 degrees, with 10 degrees
discretization). This was necessary because of the lateral asymmetry of the hull and the
propellers distribution (see Figure 6 - the scale was omitted for a matter of confidential issues).
October 11-12, 2011
Page 6
BGL1 - Experimental Current Coefficients - Captive Model

2.5
Cy
2.0
10*Cx
10*Cm
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
0
90
180
270
360
Current Related Heading (degrees)
Figure 6 Barge drag coefficients

The wind tunnel tests were done to obtain the wind coefficients and to verify the boom influence
in the barge loads. Tests were conducted with three different angles of the boom as well as
variations in the static angles of trim and rolling of the barge. These tests were performed in the
IPT's Boundary Layer Wind Tunnel, on a scale of 1:100. Figure 7 shows the barge model in the
wind tunnel and a qualitative result of the lateral force coefficient for three different boom angles.
Figure 7 (left) Wind-tunnel model test setup, (right) Lateral wind force coefficient
Propeller open-water calibration test was carried out, in order to define the relation between the
shaft rotation (rpm) and delivered thrust (N). The test set-up is presented in the Figure 8(a). The
propeller is installed in the end of a bar; that is rigidly connected to a bridge outside the tank.
Strain-gages are used to measure the deformation of the bar when rotation is imposed to the
propeller. Figure 8(b) shows a detail of the model-scale propeller and Figure 8(c) the stern part of
the model of the barge. This model was used in the validation tests that will be detailed in the
next section.
(a)
(b)
(c)
Figure 8 (a)Open water thruster calibration test (b)Azimuth thruster (c)DP barge model
October 11-12, 2011
Page 7
2.2 Validation Model-Scale and Real-Scale Tests

A wide range of small-scale are being executed DP barge (at the time this paper was written), in
order to verify the assessment of the DP system and the effects of the hydrodynamic coupling
between the barge and the platform. The same DP software used in the TPN simulator is
employed for controlling the experimental model. A brief description of the tests and the main
objectives of each one are given below:
1) DP gain calibration tests: step variations in the set-points (Figure 9a) are imposed to the
vessel, in order to verify the DP control gains. Important parameters of the step response (such
as overshoot and settling time) are used to a fine adjustment of Proportional and Derivative (PD)
control gains. No environmental condition is considered.
2) Stationkeeping tests: the DP barge model is subjected to a typical combination of waves and
wind (Figure 9b), for several heading angles, and the footprint and DP forces of the vessel are
measured. The idea is to do some verifications related to the capability plots that were obtained
by numerical computation.
3) Crane operation close to a VLCC: the model of a spread moored VLCC is installed close to the
DP barge, and a relative positioning system is used to measure the distance between both
vessels. This information is used by the DP system to control the thrusters. Two relative
configurations will be considered (parallel and perpendicular - Figure 9c). The relative motion of
the vessels, first order oscillations, DP forces and barge footprint are measured for representative
waves. The objective is to verify the hydrodynamic interference between the hulls, concerning
first order motion and drift forces.
Horizontal equivalent
mooring
Spread Moored VLCC
DP Barge Pos. 1
Surge
Sway
Yaw
Several wave
headings
Pos 2
(a)
(c)
(b)
Figure 9 (a)DP calibration test (b)Stationkeeping test (c) Crane operation test
Commissioning sea trials have been executed in a location close to the shore. Current meters
and anemometers were used to measure the environmental conditions. All DP parameters
(estimated thrust, power and azimuth angles) and barge positioning were also logged. An
independent 6 degree of freedom MRU will be used to measure the first order motion in a specific
point of the barge, and an offline filtering algorithm is used to estimate the acceleration in the tip
of the crane. DGPS position obtained from DP system is be logged. Figure 10 shows an example
of the logged data during a test of heading change, with 2 thrusters inactive. The tests have been
finished just before this paper has been concluded, so the analyses of the results have not been
done yet. The same environmental condition will be reproduced in the simulator and DP
parameters (average power, standard deviation, footprint) will be compared
October 11-12, 2011
Page 8
Thruster 1 (kN)
30
140
20
120
10
0
100
200
400
600
800
1000
1200
800
1000
1200
800
1000
1200
800
1000
1200
800
1000
1200
800
1000
1200
Heading (deg)
Thruster 2 (kN)
30
80
20
10
60
200
400
40
600
Thruster 3 (kN)
1
20
0
0
200
400
600
Time(s)
800
1000
1200
-1
200
400
600
Thruster 4 (kN)
40
1
20
0.8
0
0.6
200
400
600
Heave (m)
Thruster 5 (kN)
0.4
0.2
-1
200
400
600
Thruster 6 (kN)
-0.2
60
-0.4
40
-0.6
20
-0.8
0
0
200
400
600
Time(s)
800
1000
1200
200
400
600
Time(s)
Figure 10 Real scale monitoring time series

2.3 Numerical model and first analyses
Barge - FPSO
The first operation is the replacement of an equipment from the deck of a turret moored FPSO.
The distance between both vessels during the operation will be approximately 20m. Two auxiliary
cables are used in the operation (Figure 11). The numerical model was used to define several
operational parameters, including the safe relative position between FPSO and barge (side of the
FPSO, distance and angle) and the maximum environmental condition. Several operational
parameters were evaluated by means of the simulations: cable traction, tip-crane acceleration,
load motion, DP utilization and barge motion. For this analysis, the hydrodynamic interference
between the vessel is considered.
Lifting Cable
Equipment
Auxiliary Cables
Figure 11 Barge FPSO operation
October 11-12, 2011
Page 9
As an example, for the intermediate loading condition of the FPSO, the error DP control point
footprint is presented in the Figure 12a, considering significant wave height of 2.0m. Since the
FPSO-barge are heading to the resultant of the environmental forces, the mean DP utilization is
reduced, as shown in Figure 12b. Finally, the tip-crane acceleration (in gs) is presented in Figure
12c, for all wave headings. It can be obviously verified that it is preferable to operate in the
protected side of the FPSO. A comprehensive study about this operation is given in Vieira et al.
(2011).
X vs Y - BGL, configurao Transversal, FPSO Cheio, 16m de cabo (ec 2)
8
270
Medio P Hs=2m com swell

100
2,5
300
90
240
2
1,5
330
70
Mean Utilization (%)
Deslocamento em Y (m)
80
210
1
60
0,5
50
40
180
30
20
-2
10
0
-4
-50
-45
-40
Deslocamento em X (m)
-35
30
1
150
Thruster
60
(a)
(b)
120
90
(c)
Figure 12 - (a) Footprint of DP control point; (b) DP mean power utilization; (c) Tip-crane
acceleration
Barge - Fixed Platform
The second operation evaluate is the replacement of an equipment from the plant of the fixed
platform. The distance between both vessels during the operation will be approximately 16m.
Three auxiliary cables are used in the operation (Figure 13). Due to the truss structure of the
platform, there is a smaller hydrodynamic interference between the vessels.
Figure 13 - Barge Fixed platform operation
The Figure 15a shows the DP positioning error for the condition presented above. Comparing
with the error of FPSO operation (Figure 12a), one can verify that the hydrodynamic interference
of the FPSO on the barge plays an important role in this parameter, inducing a larger motion at
the barge. Figure 15b shows the mean DP utilization, larger than in the case of the FPSO since
the barge is not heading to the environmental resultant force. Finally, Figure 15c shows the
possible wave-headings that the barge can be kept in order to guarantee acceptable tip-crane
acceleration (smaller than 0.15g). Again, this kind of numerical analysis was used for defining
safe relative position of the barge and the maximum environmental condition.
October 11-12, 2011
Page 10
X vs Y - BGL, posio 1 (ec 2)

54
100
53
90
52
70
Deslocamento em Y (m)
80
51
50
49
48
60
50
40
47
30
46
20
10
45
44
-2
2
4
Deslocamento em X (m)
Thruster
(c)
(b)
(a)
Figure 14 Figure 15 - (a) Footprint of DP control point; (b) DP mean power utilization; (c) Tip-crane
acceleration
3. Case Study 2: DP Drilling Vessel

The second case is related to a DP drilling vessel operating under failure conditions. An incident
report is used to reproduce in the simulator the same conditions verified in a real drift and
disconnection event. The results from simulator have indicated the same overall behavior verified
in the real situation as will be detailed in the present section.
The vessel main dimensions are indicated in the Table 1. The vessel is equipped with 17
propellers/thrusters, as indicated in Figure 16.
Table 1 DP Drilling Ship main characteristics

Characteristic
Length overall
Breadth
Operating Draft
Depth
Displacement
16/17
15
11/14
10
3/6
Value
168 m
26 m
7.6 m
13 m
25.000 MT
2 1
Propeller /
Thruster
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Type
Power (kW)
Fixed Retract.
600
Fixed Retract.
600
Fixed Retract.
600
Fixed Retract.
600
Fixed Retract.
600
Fixed Retract.
600
Azimuth Retract.
1500
Azimuth Retract.
1500
Azimuth Retract.
1500
Azimuth Retract.
1500
Fixed Retract.
600
Fixed Retract.
600
Fixed Retract.
600
Fixed Retract.
600
Skeg Thruster
600
Propeller
3600
Propeller
3600
Figure 16 DP drilling vessel thrusters
October 11-12, 2011
Page 11
The environmental condition during the failure event is show in Figure 17 (left). The wind direction
and speed was obtained from the DP anemometer, and the current is estimated from the DP
system, by means of the Kalman Filter algorithm. It must be stressed that this "current" is in fact a
combination of all unmeasured external forces action on the vessel. So, it is indeed the resultant
from current and waves.
o
In the initial situation, the vessel heading was 255 , but the DP operator decided to change it to
o
120 . Heading limitations imposed by the angle of the auxiliary lines (kill and choke) did not allow
the counter-clockwise rotation. So, the heading change was done in the clockwise direction, as
o
indicated in the Figure 17. However, in an intermediate heading (320 ), thruster 8 was switched
off due to electric current limitation, and a position drift was verified. A fast emergency
disconnection operation was then executed.
DP Current
45o ; 3.4 knots
Initial Position
255o
N
Drift Position
320o
Final Desired Position

120o
Wind
120o ; 30 knots
Figure 17 (left) Environmental condition and desired position ; (right) intermediate failure
position
The mean utilization (just before the drifting off) of the thrusters (excluding the propellers) was
about 80% and of the propellers was about 10%. Furthermore, during the analysis of this event, it
was verified that the rotations of thrusters 5, 11 and 15 were inverted. It affected the DP holding
capacity of the vessel, since those thrusters were acting against the required force.
The same situation was reproduced in the TPN simulator. Figure 18 shows the mean utilization
o
just before the failure of the thruster 8, considering that the vessel heading is 320 . The thrusters
with inverted rotation were also considered and are indicated in the figure. The comparison
between the simulation results and the real-incident observations are summarized below:
- the average utilization of the thrusters 1 to 15 (excluding the propellers) obtained in the
simulation is 76%, very close to the value of 80% given in the incident report;
- the azimuth thrusters (7 to 10) are in fact over-demanded, what should cause the switch-off of
thruster 8 in the real incident;
- the mean utilization of the propellers (16 and 17) obtained in the simulation is 23%, larger than
the 10% given in the incident report. Since both values indicated that those propellers are not
over-demanded, this discrepancy is not relevant in the present analysis. Small differences in the
longitudinal forces estimation may be responsible for this difference.
October 11-12, 2011
Page 12
110
100
90
Inverted rotation
80
70
60
50
40
30
20
10
0
10
12
14
16
18
Thruster
Figure 18 Mean thruster utilization just before the failure of thruster 8 (Simulation)
The drifting of the DP vessel was also verified in the time-domain simulation, as indicated in the
Figure 19. Just after the failure of thruster 8, the vessel cannot keep position and a large offset
from the required position was verified.
150
Offset (m)
100
50
0
3500
Failure of
thruster 8
4000
4500
5000
5500
6000
Time(s)
6500
7000
7500
8000
Figure 19 Offset of the DP drilling vessel after the failure of thruster 8 (Simulation)
The previous analysis indicated a very good qualitative and quantitative agreement between the
numerical simulation results and the real-incident observations. This may be considered as a
validation of the simulator. After that, the simulator may be used for evaluating a more adequate
angle for the BOP, considering the typical environmental conditions of the Brazilian offshore oil
fields and common variations of wind conditions in those fields.
4. Case Study 3: DP Offloading Operation

Since 1994 Petrobras decided to apply DP Shuttle Tankers (DP-ST) for oil-offloading, and due to
market and cost constraints, the first solution was the conversion of old conventional tankers. An
Enhanced Class-1 DP was defined as an acceptable standard, with redundancy in some
important systems. Despite this original strategy to speed up the DP Shuttle Program
October 11-12, 2011
Page 13
implementation, Petrobras has now defined the DP Class-2 as a minimum for the new vessels
contracted or to be built from now on (Sphaier et al., 2009).
With this constantly evolution scenario, Petrobras has started cooperative programs with Brazilian
scientific community, in order to carry out experimental and numerical studies to support
important decisions about regulation, design and operational requirements for the offloading
systems and new vessels. Such works resulted in important design and analysis methodologies
as well as design tools, such as time-domain simulators and experimental facilities.
This section will present the validation of the numerical time-domain simulation, by means of
small-scale experiments. After that, the simulator was applied to several analyses, and two
examples will be exposed. The first one is a comparative station-keeping and failure analysis of
the DP1-Enhanced / DP-2 STs considering typical environmental conditions of Brazilian oil fields.
The second analysis is a comprehensive downtime evaluation procedure that can be used as an
important design parameter for DP holding power requirement.
4.1 Small-scale experimental validation of the numerical simulator
The experimental tests were conducted in the Laboceano Laboratory (Rio de Janeiro, Brazil), at
the main ocean basin, in order to validate the numerical models of DP-ST. A detailed description
of those results is given in Tannuri et al., (2010), and a brief summary of main results are
exposed in the present sub-section.
A typical Suezmax DP2-ST is considered in the present work. The 1:70 model-scale main
properties in the two loading conditions considered in the analysis are given in Table 2.
Table 2. Model main characteristics
Main Dimensions
Total Length (mm)
Length betw. perp. Lpp (mm)
Beam (mm)
Depth (mm)
Draft (mm)
Total Displacement (kg)
Full
Ballasted
3843
3686
657
349
250
115
498
215
An illustrative picture of the test is presented in Figure 20(a). Wave and wind actions were
considered. The control architecture (software, hardware, measurement system, integration) of
the test was developed in a current cooperation project between Petrobras and Brazilian
Universities (Morishita et al., 2009).
The coordinate system and angle definitions are illustrated in Figure 20(b). The vessel heading
o
angle () is 180 when the vessel is heading to the wave generator. A positive heading angle is
o
illustrated in the figure. Wave and wind directions are 180 when they are directed toward the
model. The origin of XY tank coordinate system is also illustrated. During the experiments the
midship section of the vessel is controlled around X=20m; Y=15m, equivalent to the central point
of the tank. The vessel is equipped with 2 tunnel-thrusters, 2 azimuth thrusters and main
propeller. The rudder is not used in DP mode. The positions of the thusters (related to mid-ship
section) and the maximum thrusts are shown in Figure 20(c).
#5 Tunnel Bow
Fmax=0.71N
Fmin=-0.71N
X=1810mm
#1 - Main
Fmax=3.51N
Fmin=- 1.70N
X= -1750mm
1:70 model
#2 Tunnel Stern
Fmax=+0.53N
Fmin=-0.53N
X= -1440mm
(a)
(b)
#3 Azim Stern
Fmax=+1.01N
X= -1072mm
#4 Azim Bow
Fmax=+1.01N
X=1637mm
(c)
Figure 20 (a) Experimental test; (b) Coordinate system definition; (c) Thrusters layout
October 11-12, 2011
Page 14
The environmental conditions considered in the tests are presented in Table 3. The COL2
condition is equivalent to the limiting environmental condition that the offloading operation can be
executed in Brazilian waters (Tannuri et al., 2009a). The COL1 and COL3 are equivalent but
coming from different directions (with waves slightly weaker than COL2).
Table 3. Environmental conditions
Condition
Wind
Wave
COL1
2.4m/s 180o Hs=0.036m; Tp=0.95s 180o
COL2
2.4m/s 180o Hs=0.050m; Tp=1.08s 180o
COL3
2.4m/s 225o Hs=0.036m; Tp=0.95s 225o
The station-keeping test with the full-loaded model, =180 and COL1 condition (bow incidence)
is presented in Figure 21. Quite acceptable agreement is verified for vessel motion and thrusters
forces. The mean utilization for this case and other full loaded cases is presented in Figure 22.
For both cases a very good agreement between simulation and experimental results could be
asserted.
o
F(N) Main Prop. Mean=0.38N Std=0.0313 N
X(m) Std=0.00381m
X(m) Std=0.00429m
20.02
20.02
20.01
20.01
X Setpoint (m)
X Position (m)
20
20
1000
19.98
1000
X Setpoint (m)
X Position (m)
1100
1200
1300
1400
1500
Time(s)
1600
1700
1800
1900
2000
19.98
1000
1100
1200
1300
1400
1500
Time(s)
1600
1700
1800
1900
2000
1200
1300
Y Setpoint (m)
Y Position (m)
15.01
15
15
1100
1200
1300
1400
1500
Time(s)
1600
1700
1800
1900
2000
14.98
1000
1200
1300
Heading (o) Std=0.0482o
1400
1500
Time(s)
1600
1700
1800
1900
2000
Heading ( ) Std=0.0459
Heading Setpoint (o)

Heading Angle (o)
1300
1400
1500
Time(s)
1600
1700
1800
1900
2000
179.8
1000
1000
1100
1200
1100
1200
1300
1400
1500
1600
1700
1800
1900
0.5
1200
1300
1400
1500
1600
1700
Simulation
1200
1300
1400
1500
Time(s)
1600
1700
1800
1900
1100
1200
1800
1900
0
1000
1100
1200
Mean=0.414N Std=0.0577 N
2000
1000
1300
1400
1300
1400
F(N) Azim. Bow
0
1000
1400
1500
1600
1700
1800
1900
2000
1500
1600
1700
1800
1900
2000
1800
1900
2000
1800
1900
2000
1900
2000
F(N) Azim. Stern Mean=0.478N Std=0.0576 N

1
0.5
1500
1600
1700
Mean=0.464N Std=0.0575 N
1
0.5
1100
1200
1300
1400
1500
1600
1700
1800
1900
0
1000
1100
1200
1300
1400
1500
1600
1700
F(N) Tunnel Bow Mean=0.0534N Std=0.0658 N

0.5
1100
-0.5
1000
F(N) Azim. Stern Mean=0.42N Std=0.0581 N

1
1100
1300
F(N) Tunnel Stern Mean=0.132N Std=0.0835 N
-0.5
Heading Angle (o)

1200
1900
0.5
179.9
Heading Setpoint (o)
1100
1800
F(N) Tunnel Bow Mean=0.0923N Std=0.114 N
180
179.9
1700
0.5
180.1
180
1600
0.5
F(N) Azim. Bow
1100
180.1
1500
0
1000
14.99
14.99
1400
0.5
-0.5
1000
Y(m) Std=0.000727m
15.02
Y Setpoint (m)
Y Position (m)
15.01
179.8
1000
1100
F(N) Tunnel Stern Mean=0.1N Std=0.128 N
Y(m) Std=0.00212m
15.02
14.98
1000
3
2
1
0
-1
19.99
19.99
F(N) Main Prop. Mean=0.46N Std=0.0582 N
3
2
1
0
-1
-0.5
1100
1200
1300
1400
1500
Time(s)
1600
1700
1800
1900
1000
1100
1200
Simulation
Experiment
1300
1400
1500
Time(s)
1600
1700
1800
Experiment
Figure 21 Full, COL1
COL2 =180o
COL1
COL3 =240o
Figure 22 Mean Thrusters utilization

Several other conditions are presented in Tannuri et al., (2010). The validation of the numerical
simulator has also been done considering full scale measurements of a DP offloading operation.
Good agreement between numerical results and full-scale measurements was observed, and the
main results of such campaign can be found in Tannuri et al. (2009c) and Saad et al. (2009).
4.2 Comparative Station-keeping and Failure Analysis

A comprehensive comparative station-keeping and failure analysis among DP1-Enhanced and
DP-2 shuttle tankers were done, considering typical environmental conditions of Brazilian oil
October 11-12, 2011
Page 15
fields (Campos Basin). A spread-moored FPSO heading to SW is considered, and the bow
operation is simulated for the conditions shown in Figure 23.
Correnteza
Current
S SS
Correnteza
0,7m/s
0,7m/s
Current
S S
Correnteza
0,7m/s
VentoNE
NE
Wind
Vento
12m/s
12m/s
OndaNE
NE
Onda
Wave
NE
Hs=1,98mTp=6.8s
Tp=6.8s
Hs=1,98m
VentoSE
SE
Wind
10.5m/s
Swell SE
Hs=4,0m Tp=14s
A partir de t=7000s
Optional
Onda SE
Wave
SE
Hs=1,97m Tp=11.2s
Figure 23 Environmental conditions (left) A ; (right) B

Three different Suezmax-class DP-STs were considered, indicated in the Table 4. The DP1Enhanced is a typical configuration of the first generation of DP-ST operating in Brazilian waters.
The DP2a is the configuration of the second generation of DP vessels. Finally, DP2b is a new
proposal for vessels to be operated in pre-salt fields (Santos Basin).
Table 4. Thrusters layout

DP1-Enhanced
Propeller
1- Tunnel Thruster Bow
2- Azimuth Stern Thruster
3- Azimuth Bow Thruster
4- Main Propeller
DP2a - vessels being constructed

Power
2200kW
2200kW
2200kW
18.881kW
Propeller
2- Azimuth Thruster Bow
3- Tunnel Thruster Stern
4- Azimuth Thruster Stern
6- Main Propeller
Power
2200kW
2200kW
1200kW
2000kW
14.280kW
DP2b - Alternative for pre-salt fields

Propeller
4- Azimuth Thruster Stern
5- Tunnel Thruster Stern
6- Main Propeller
Power
2200kW
2200kW
2200kW
2200kW
1400kW
16.860kW
Figure 24 shows the results for the condition B, considering the failure of the bow azimuth
thruster. The DP1-Enhanced ST cannot keep position after the failure, with saturation of thrusters
1 and 2. The DP2a could hold position even after the failure with a weathervane control, but a
very large utilization and saturation of the remaining thrusters was verified, in order to
compensate for the failure. DP2b was able to keep position aligned to the FPSO axis, and due to
the redundancy of the thrusters, no saturation was verified after the failure. Similar analysis was
done for other conditions, in order to support technical decisions concerning DP power capacity
for new tankers.
4.2 Downtime Analysis

Finally, the simulator was used for a comparative analysis of DP oil carriers concerning the
downtime for offloading operations in a spread-moored FPSO. The procedure consisted in
evaluating (by means of a time-domain simulation or static analysis) if the ST can be safely kept
position for a comprehensive set of environmental conditions. The Figure 25 shows the allowable
o
o
positions for a typical SMS FPSO, since the angular safety zone is defined as -45 / +60 .
October 11-12, 2011
Page 16
Delivered Force per Thruster (kN)
100
100
-100
-100
Thrus.1
200
0
-200
5000
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
Time (s)
8000
8500
9000
9500
400
300
-200
-300
200
Thrus.2
Y (m)
Y (m)
-200
100
0
-300
-100
5000
-400
-400
400
300
-500
-600
-500
-400
-300
-200
-100
100
-600
-500
200
X (m)
Thrus.3
-500
-400
-300
-200
-100
100
200
100
0
200
X (m)
-100
5000
2000
Before failure of azimuth bow thruster
Thrus.4
After failure of azimuth bow thruster

1000
0
-1000
5000
100
100
-100
-100
-200
-200
Thrus.1
200
0
-200
5000
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
Time (s)
8000
8500
9000
9500
400
200
100
0
Y (m)
Y (m)
Thrus.2
300
-300
5000
-300
-400
Thrus.3
100
-400
-500
0
-100
-200
5000
-500
-600
-500
-400
-300
-200
-100
100
-600
-500
200
-400
-300
-200
X (m)
-100
100
200
X (m)
Thrus.4
400
200
0
5000

Thrus.5
1000
500
0
5000
100
100
Thrus.1
200
0
-200
5000
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
8000
8500
9000
9500
5500
6000
6500
7000
7500
Time (s)
8000
8500
9000
9500
-200
-200
200
0
5000
Y (m)
-100
400
-300
-300
-400
-400
Thrus.3
Y (m)
-100
Thrus.2
400
200
0
5000
-500
-600
-500
Thrus.4
400
-500
-400
-300
-200
-100
100
200
-600
-500
-400
-300
-200
-100
100
0
5000
200
X (m)
Thrus.5
X (m)
200
100
0
-100
Thrus.6
5000
1000
0
-1000
5000
Figure 24 Condition B ; Full-loaded ST, (top) DP1-Enhanced; (Middle) DP2a ; (Down) DP2b
October 11-12, 2011
Page 17
60
45
N
Stern Operation
FPSO
60
Bow Operation
45
Figure 25 Relative positions FPSO / ST

Two important questions must be previously studied, related to the definition of the environmental
conditions that represent the long-term state of the location and the validation of a static approach
for calculation the station-keeping ability, as an alternative to the time-consuming dynamic
analysis. So, a procedure for defining the comprehensive set of environmental conditions was
established. Since the number of environmental conditions may be extremely large (in the
present analysis, 200,000), a static power calculation must be used. So, some time-domain
simulations with automated post-processing were done and compared to static calculation, in
order to define the dynamic margin that must be included in the static calculation to obtain similar
results. After that, the static calculation can be used for estimating the downtime for several DPlayouts, relative positions between ST and FPSO and FPSO heading. The downtime is an
important design parameter for ST-DP definition and some illustrative results will be presented.
Comparison between static and dynamic analysis
In order to validate the static calculation of power holding capacity, a comparison between timedomain simulation results and static calculation was done for the DP1-Enhanced shuttle tanker,
under 25.550 environmental conditions. The static calculation does not consider slow drift forces,
transient motions and wind gusts, so a dynamic margin must be included in order to account for
them. Normally, it is used 20% margin.
However, the present analysis showed that for offloading analysis, a 15% margin is acceptable,
since the ST may present large motions around the desired set-point. Figure 26 shows the results
for three different ST headings (related to North) and two loading conditions. It can be verified
that for all cases, the static analysis with margin of 20% over-estimates the downtime. The static
analysis with 15% margin also overestimates, but the difference comparing to the dynamic results
are quite reduced. It was also verified the if a 10% margin is used, the downtime may be underestimated for same cases, what indicates that some dynamic factors are in fact between 10% and
15%. So, the value of 15% was used for all the next analysis. More complete description of this
analysis is presented in Tannuri et al. (2009b).
Full Loaded Condition
Ballasted Condition
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
42
62
82
Static 20%
Static 15%
Dynamic
42
ST Heading (deg)
62
82
ST Heading (deg)
Figure 26 Downtime estimation for 25.550 typical conditions in Campos Basin (DP1Enhanced)
October 11-12, 2011
Page 18
Definition of environmental conditions

Another important issue for estimating the downtime is related to definition of a representative set
of environmental conditions. In our case, a simultaneous register of waves and wind during 11year monitoring campaign was available, on a daily basis. However, current was not measured in
this campaign, and the Metocean data should then be used. Considering the hypothesis of
statistical independence between wave/wind and current, a set of 68 representative currents
(speed/direction) was used. The occurrence of currents was also obtained from the Metocean,
and each of them was combined with each wave/wind register.
A sensitivity analysis was done to verify if a smaller number of years could be used, in order to
speed up the downtime estimation. The Figure 27 shows the downtime estimated for the DP1Enhanced ST, for stern or bow operation in the FPSO. One can verify that there is a good
estimation for 8 years or more, so it can be considered that an acceptable accuracy is obtained
using 8-year monitoring campaign.
10.00%
9.00%
8.00%
7.00%
6.00%
Full Bow
5.00%
4.00%
Full Stern
3.00%
Ballast Bow
2.00%
Ballast Stern
1.00%
0.00%
1
10
11
Number of years
Figure 27 Sensitivity analysis considering the number of years

So, the downtime is then estimated by the calculation of DP holding-capacity for each heading
considering (8year x 365 days) waves/wind registers, each of them combined to the 68 possible
currents states (total of approximately 200,000 conditions).
Downtime Results
For the DP1-Enhanced operating in a typical SMS-FPSO in Campos Basin, the downtime
estimation procedure was applied. For each one of the 200,000 conditions, the ST was verified if
it could keep position for any heading inside the operational zone. If there is any possible
position, the occurrence is considered as uptime, otherwise it is considered as downtime. For
example, for the environmental conditions of Figure 23, the next figure shows that three
conditions can be considered as uptime, since there are safe-positions inside the operational
sector.
Figure 28 Safe positions for bow-operation, DP1-Enhanced, 3 typical conditions
October 11-12, 2011
Page 19
The final results of Table 5 were obtained. It can be seen that the bow operation is preferable,
since for Campos Basin the current comes preferable from North direction.
Table 5. Downtime results for DP1-Enhanced ST
Type of offloading station active
Downtime
Total (Bow or stern operation)
5.3%
Stern Operation Only
62.0%
Bow Operation Only
21.5%
Other important analyses using the downtime estimation were done using this procedure, but the
results could not be publically exposed here. Those analyses included:
-
Comparative analysis of downtime for the DP1-Enhanced or DP2 shuttle tankers

Downtime analysis for the new pre-salt oil-fields in Santos basin
Downtime analysis for several headings of SMS-FPSO in Santos Basin, in order to aid
the design of the mooring system for new FPSOs to be installed there
Potential advantages (in terms of uptime) considering the augmentation of the safety
zone.
5. CONCLUSIONS
This paper illustrated the application of numerical simulation tools for aiding and supporting
decisions concerning real applications of offshore DP operations. Three cases have been
considered: a DP crane barge, a DP drilling vessel and DP shuttle tankers.
6. ACKNOWLEDGEMENT
The authors thanks to Petrobras and Transpetro for the constant support for the technological
development related to DP applications. The first author thanks to National Counsel of
Technological and Scientific Development (CNPq) for the research grant (302544/2010-0) and
The Brazilian Innovation Agency (FINEP) for the financial support of part of this research
(01.10.0748.00).
7. REFERENCES
ARANHA, J.A.P., A Formula for Wave Damping in the Drift of a Floating Body, Journal of Fluid
Mechanics, vol. 272, pp.147-155, 1994.
FALTINSEN, O.M., Sea Loads on Ships and Offshore Structures, Cambridge, Cambridge
University Press, England, 1990.
MORISHITA, H.M. ; TANNURI, E. A. ; SAAD, A. C. ; SPHAIER, S. H. ; LAGO, G.A. ;
MORATELLI JR., L. . Laboratory Facilities for Dynamic Positioning System. In: 8th Conference on
Maneuvering and Control of Marine Craft (MCMC'2009), Brazil, 2009.
OCIMF, Predictions of wind and current loads on VLCCs, Oil Companies International Marine
Forum, 1994.
PINKSTER, J.A., Low Frequency Second Order Wave Exciting Forces on Floating Structures,
PhD Thesis, Delft University of Technology, The Netherlands, 1988.
RAMPAZZO, F. ; SILVA, J.L.B. ; VIEIRA, D. P. ; PACIFICO, A. L. ; MORATELLI JR., L. ;
TANNURI, E. A. . Numerical & Experimental tools for offshore DP operations. In: ASME 30th
October 11-12, 2011
Page 20
International Conference on Ocean, Offshore and Arctic Engineering, OMAE2011, Roterdam,

The Netherlands, 2011.
SAAD, A.C., VILAIN, L., LOUREIRO, R., BRANDO, R., LOPES, C., GIOPPO, H., Installation
and operation of the mono-column FPSO Sevan Piranema in Brazilian waters, Proceedings of
OTC 2009, 2009.
SIMOS, A.N. ; TANNURI, E. A. ; PESCE, C.P. ; ARANHA, J. A. P. . A quasi-explicit
hydrodynamic model for the dynamic analysis of a moored FPSO under current action. Journal of
Ship Research, v. 45, n. 4, p. 289-301, 2001.
SPHAIER, S. H. ; CORREA DA SILVA, S.H.S. ; CORREIA, C. A. R. ; MORISHITA, H.M. ;
TANNURI, E. A. . Evolution of Crude Oil Offshore Loading in Brazilian Waters. In: 8th Conference
on Maneuvering and Control of Marine Craft (MCMC'2009), Brazil, 2009.
TANNURI, E. A., SAAD, A. C., CORREA DA SILVA, S.H.S., MATOS, V. L. F., MORISHITA,
H.M., SIMOS, A.N., SPHAIER, S. H. . DP Assisted Offloading Operations in Brazilian Waters. In:
Dynamic Positioning Conference (DP-Conference), Houston, USA, 2009a.
TANNURI, E. A. ; SILVA, J.L.B. ; OSHIRO, A.T. ; AZEVEDO JR., P. C. . Design and Analysis of
Dynamic Positioning Systems: Comparison Between Static and Dynamic Approaches. In: 10th
International Marine Design Conference, 2009, Trondheim. Proceedings of 10th International
Marine Design Conference, 2009b.
TANNURI, E. A.; SAAD, A. C. ; MORISHITA, H.M. . Offloading Operation with a DP Shuttle
Tanker: Comparison Between Full Scale Measurements and Numerical Simulation Results. In:
8th Conference on Maneuvering and Control of Marine Craft (MCMC'2009), Brazil, 2009c.
TANNURI, E.A. and MORISHITA, H.M., Experimental and Numerical Evaluation of a Typical
Dynamic Positioning System, Applied Ocean Research, vol. 28 pp. 133-146, 2006.
TANNURI, E. A.; MORISHITA, H.M.; VILAA, R.; SAAD, A. C. . Numerical and experimental
analysis of a typical DP shuttle tanker operating in Brazilian waters. In: 8th IFAC Conference on
Control Applications in Marine Systems, CAMS 2010, Rostock-Warnemnde, Gernamy, 2010.
VIEIRA, D. P. ; MALTA, E.B. ; RAMPAZZO, F. ; SILVA, J.L.B. ; TANNURI, E. A. . Effects of
coupled hydrodynamic in the performance of a DP barge operating close to a FPSO. In: ASME
30th International Conference on Ocean, Offshore and Arctic Engineering, OMAE2011,
Roterdam, The Netherlands, 2011.
WAMIT, WAMIT User Manual Versions 6.0, 6.0PC, 5.3S, WAMIT Inc., MA, USA, 2000.
October 11-12, 2011
Page 21

Newapps Tannuri

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Newapps Tannuri

Caricato da

Copyright:

Formati disponibili

Return to Session Directory

Name of the Paper

DYNAMIC POSITIONING CONFERENCE

NEW APPLICATIONS SESSION

Utilization of numerical simulation tools for aiding

(1) University of So Paulo; (2) Petrleo Brasileiro S/A - Petrobras;

Tannuri, E.A., et al.

New Applications Session

Utilization of numerical simulation...

MTS Dynamic Positioning Conference

October 11-12, 2011

Tannuri, E.A., et al.

New Applications Session

Utilization of numerical simulation...

1. Time Domains Simulation Code

Floating body main parameters (dimensions, mass matrix, etc.);

Figure 1 TPN block diagram

MTS Dynamic Positioning Conference

October 11-12, 2011

Tannuri, E.A., et al.

New Applications Session

Utilization of numerical simulation...

Figure 2 Coordinate systems

YL T( ) x 2 L , with T( L ) sin( L ) cos( L ) 0

Wave-frequency (first-order) motions ( X H , YH , H ) are evaluated by means of transfer functions

MTS Dynamic Positioning Conference

October 11-12, 2011

Tannuri, E.A., et al.

New Applications Session

Utilization of numerical simulation...

2. Case Study 1: DP Crane Barge

Figure 3 DP crane barge operation (left) FPSO ; (right) Jack-up platform

MTS Dynamic Positioning Conference

October 11-12, 2011

Tannuri, E.A., et al.

New Applications Session

Utilization of numerical simulation...

Figure 4 Flowchart of the project for evaluating DP Crane Barge operation

MTS Dynamic Positioning Conference

October 11-12, 2011

Tannuri, E.A., et al.

New Applications Session

Utilization of numerical simulation...

BGL1 - Experimental Current Coefficients - Captive Model

Current Related Heading (degrees)

Figure 6 Barge drag coefficients

MTS Dynamic Positioning Conference

October 11-12, 2011

Tannuri, E.A., et al.

New Applications Session

Utilization of numerical simulation...

2.2 Validation Model-Scale and Real-Scale Tests

Spread Moored VLCC

MTS Dynamic Positioning Conference

October 11-12, 2011

Tannuri, E.A., et al.

New Applications Session

Utilization of numerical simulation...

Figure 10 Real scale monitoring time series

Figure 11 Barge FPSO operation

MTS Dynamic Positioning Conference

October 11-12, 2011

Tannuri, E.A., et al.