You are on page 1of 8

1 Copyright 2013 by ASME

Proceedings of the ASME 2013 32nd International Conference on Ocean, Offshore and Arctic Engineering
OMAE2013
June 9-14, 2013, Nantes, France

OMAE2013-11569
COUPLED TRANSIENT CFD AND DIFFRACTION MODELING FOR INSTALLATION
OF SUBSEA EQUIPMENT/STRUCTURES IN SPLASH ZONE


David Jia
Technip
Houston, Texas, USA
Madhusuden Agrawal
ANSYS
Houston, Texas, USA





ABSTRACT
In development of deep water oil and gas fields,
successfully and economically installing subsea equipment and
structure is critically important. This paper presents a state-of-
the-art methodology for predicting the motions and loads of
subsea equipment/structure during such operations basing on
time domain simulations of the combined installation vessel
and subsea equipment/structure. The time domain diffraction
simulation of the moving lifting vessel is coupled with
multiphase CFD simulation of subsea equipment/structure in
splash zone. Transient CFD model with rigid body motion for
the equipment/structure calculates added masses, forces and
moments on the equipment/structure for diffraction analysis,
while diffraction analysis calculates linear and angular
velocities for CFD simulation. This paper has many potential
applications, such as, installation of pile, manifold, subsea tree,
PLET/PLEM, or other subsea equipment/structure. This
coupled approach has been successfully implemented on a
cylindrical structure. The results show that total load level, and
dynamics of the subsea equipment/structure due to waves in
splash zone are predicted. Current practice of installation
analysis in accordance with the recommendations from DNV-
RP-H103 [1] cannot determine in detail the wave loads either
during the passage through splash zone, or added mass and
damping when the equipment/structure is submerged. In order
to determine wave loads in detail, model tests are needed. In the
absence of tests, simplified equations or empirical formulations
have to be used to calculate/estimate these hydrodynamics
coefficients as recommended in DNV-RP-H103. Steady-state
CFD simulations on a stationary equipment/structure are
usually used to predict drag and added masses on submerged
structures. However the steady-state assumption in CFD
ignores the resonating motion of equipment/structure in
calculating hydrodynamics coefficients, which can severely
affect the accuracy of these predictions. The above methods
often give overly conservative results for allowable sea state
which results in uneconomical vessel time or inaccurate results
for installation. The methodology of this paper gives more
accurate results, and provides potentially economical vessel
time during installation. The intent of this paper is to
demonstrate the solution and methodology.

INTRODUCTION
A successful offshore installation is crucial for deep water
oil and gas field development. The installation of large subsea
equipment and structures is a challenge in offshore installation
campaigns. Extensive installation analyses are performed to
ensure sufficient clearance and accessibility, sufficient
structural capacity, sufficient stability of the equipment and
structures in the installation operations. Maximum allowable
environmental conditions, sea states of the operations, are
determined using the installation analyses. Current practice is to
do time domain analysis in accordance with the
recommendations from DNV-RP-H103. Added mass, damping
and hydrodynamic coefficients are calculated using DNV-RP-
H103. Analytic work was combined with experiment to predict
added mass and damping of structures that is submerged below
free surface [2] and [3]. Since large subsea equipment and
structures have complex geometric configurations, steady-state
CFD simulations on a stationary equipment/structure are
usually used to predict drag coefficient and added masses on
submerged structures in current practice of installation
analyses. These predicted hydrodynamic parameters are used in
time domain analysis.
The current practice has the inaccuracy inherited from assumed
or calculated hydrodynamic properties from stead state analyses
as recommended by DNV-RP-H103.
2 Copyright 2013 by ASME
Offshore installation of subsea equipment/structures
typically consists of the following phases: offloading from
offshore supply vessel alongside the installation vessel,
upending on the back deck of the installation vessel, lowering
through the splash zone, transferring from vessel crane to A&R
wire, lowering through water column, and landing on the sea
bed. Lowering through the splash zone is the most challenging
phase to analyze in installation analyses. This phase is usually
the governing phase for the offshore installation. Therefore, it is
critical to accurately analyze and simulate this phase in offshore
installation analyses for planning vessel & equipment assets,
and determining offshore installation weather windows and
schedule.
A research effort has been undertaken to develop a better
and accurate methodology of simulating subsea
equipment/structure lowering through the splash zone. This
paper is the result of this research effort. This paper presents the
methodology and demonstrates the solution of two-way
coupled transient CFD and time domain diffraction analysis. An
actual case installation analysis will be performed in a follow-
up paper to predict total load level, dynamics of the subsea
equipment/structure due to waves in splash zone, forces,
stresses and fatigue damage in different sea states, and give
recommendations for allowable sea states of installation.
NOMENCLATURE
A&R Abandonment and Recovery.
CFD Computational Fluid Dynamics.
COG Center of Gravity.
PLET Pipe Line End Termination.
PLEM Pipe Line End Manifold.
VOF Volume of Fluid.
METHODOLOGY OF COUPLED TRANSIENT CFD AND
TIME DOMAIN DIFFRACTION MODELLING
Automated coupling was developed to seamlessly pass
information from CFD simulation to Diffraction analysis at
every time step. User defined subroutines were written for
FLUENT software which will write forces and moments
vectors on the subsea equipment/structure geometry in a file at
every time step. This information is then read in AQWA as an
input, which will then writes out linear and angular velocities of
the subsea equipment/structure in a separate file. This file is
then read into FLUENT as an input. This two-way coupling is
performed in an automated manner at every time step.
Following sketch shows this coupling in a schematic
diagram



Figure 1: Two-way coupling diagram.

The coupling routine will output the center of gravity
vector, orientation, forces and moments on the subsea
equipment/structure at every time step. FLUENT user defined
subroutine, mainly has two functions. First function reads in the
linear and angular velocities from AQWA output file, and
assign rigid body motion to the solid zone of the subsea
equipment/structure at the beginning of the time step. The
second function calculates pressure and viscous forces as well
as moments on the body of the subsea equipment/structure and
writes them in an ASCII file at the end of the time step.
Similarly the user subroutine for AQWA has two components,
first it reads forces and moments and performs global analysis
and outputs linear and angular velocities.

VOF Transient CFD
In Computational Fluid Dynamics, conservation equations
for mass, momentum and turbulence for all phases are solved in
each control volume. In VOF model, the tracking of the
interface(s) between the phases is accomplished by the solution
of a continuity equation for the volume fraction of one (or
more) of the phases. The VOF model can model two or more
immiscible fluids by solving a single set of momentum
equations and tracking the volume fraction of each of the fluids
throughout the domain.
For the q
th
phase, this equation has the following form [4]:





(1.1)
where m
pq
is the mass transfer from phase p to phase q and m
qp

is the mass transfer from phase q to phase p. The source term
on the right-hand side of equation S

is considered zero for this


study.
A single momentum equation is solved throughout the
domain, and the resulting velocity field is shared among the
phases. The momentum equation, shown below, is dependent
on the volume fractions of all phases through the properties
and [4]:





(1.2)

Sea surface condition is categorized into air-water two
phase flow, which is suitable to use VOF method as these two
phases are immiscible. The calculation adopts finite volume
method. Second order discretization scheme was used for
higher accuracy. Volume of Fluid multiphase model with Open
Channel Boundary Condition was used in FLUENT simulation
to include wave effects. Moving Deforming Mesh Motion was
used to update solid body position of the subsea
equipment/structure at every time step based on linear and
( ) ( ) ( )
(

+ =
(

V +
c
c

=
n
p
pq pq q q q q q
q
m m S v
t
q
1
1

o
o o

( ) ( ) ( ) | | F g v v p v v v
t
T
+ + V + V V + V =
(

V +
c
c

3 Copyright 2013 by ASME
angular velocities. A non-conformal interface was created in a
spherical domain around the solid body of the subsea
equipment/structure to avoid bad skewness after remeshing the
domain. VOF approach model can model two or more
immiscible fluids by solving a single set of momentum
equations and tracking the volume fraction of each of the
immiscible fluids throughout the domain. Open channel wave
boundary conditions were used to simulate the propagation of
sea waves. A fifth order stokes wave theory was used to specify
the non-linear waves. Wavelength of 100m and wave amplitude
of 5m was used to specify the 5th order stokes wave.

Time Domain Diffraction Modeling
AQWA is a hydrodynamic radiation/diffraction code that
enables the modeling of cables and moorings. The lowering of
equipment through the splash zone is difficult for such a code
to simulate directly since the variable immersion cannot be
accounted for in the diffraction/radiation loading (would
require multiple diffraction analyses). F-K and dynamic
hydrostatic loads can be included for the variable immersion
scenarios. Drag loading is not directly included in a diffraction
based solution, neither is slamming. Because of this, it is often
simplified by modeling the equipment using Morison elements
with appropriately computed drag and added mass coefficients
(normally from CFD).
For modeling of subsea equipment installation in a
diffraction analysis, we assume the top of the line is fixed,
although this could be moving. The equipment is modeled
using a Morison element, so is very simple. We could use a
diffraction model, but it would not capture all variables for
wave loading since the diffraction and radiation effects would
vary as the body becomes immersed. Incident and variable
hydrostatic loads can be accounted for in the AQWA model, but
this was deemed insufficient accuracy for the desired
simulation, hence the coupled CFD/Hydrodynamics is used.
MODELING OF COUPLED PROBLEM
A 3D computational domain was considered for CFD
calculations. The schematic of this computational domain is
shown in Figure 2. The geometry was constructed using
ANSYS DesignModeler and the meshes were created using
ANSYS Meshing. Around 200,000 unstructured cells were used
for defining the solution domain. Finer grids were used near the
solid body and near the sea surface to capture the waves. The
computational domain was 200m in the wave direction, 400m
in the gravity direction and 100m in the transverse direction.
Figure 3 shows the picture of the mesh of the entire
computational domain on the middle vertical plane. The
computational domain consists of three different mesh zones, a
moving mesh zone at top which has a sphere mesh zone to
account for non-conformal meshing, and a deformable mesh
zone at the bottom. The sphere mesh zone contains the
cylindrical structure. The sphere mesh zone is descending into
the free surface of the sea, while the bottom mesh zone is
deformed.
Figure 4 shows enlarged view of the mesh near the
cylindrical structure on the middle vertical plane.
A refined computational mesh along with five boundary
layers was created near the cylindrical structure to properly
capture near wall effects for viscous drag. The boundary layer
details are shown in Figure 5 and Figure 6.




Figure 2: Geometry of the domain.


Figure 3: Mesh on the middle vertical plane.

Turbulence was modeled using Realizable k- model.
Standard wall functions were used at wall boundaries. A
Pressure Implicit with Splitting of Operators (PISO) scheme
was used for pressure-velocity coupling. The top surface was
modeled as pressure outlet.
4 Copyright 2013 by ASME
Coupled CFD simulation was performed for about 21
seconds of simulation which was large enough to demonstrate a
steel cylindrical structure going through the splash zone. Time
step size of 0.01 second was used for both CFD and Diffraction
analysis.


Figure 4: Enlarged view of the mesh near the
cylindrical structure geometry.


Figure 5: Zoomed-in view of the overall boundary layer
mesh near the cylindrical structure geometry.


Figure 6: Local view of the boundary layer mesh near
the cylindrical structure geometry.
NUMERICAL SIMULATION AND RESULTS
To demonstrate the methodology and solution, a steel
cylindrical structure with a length 25.33 m, diameter 4.88 m,
and mass 158 metric ton is simulated using the methodology
proposed in this paper. A total of about 21 seconds is simulated
with a time step of 0.01 second.
Figure 7 shows a series of snapshots of water volume
fraction on the middle vertical plane. The snapshots show the
motion of the cylindrical structure in the waves when it is being
lowered into the splash zone.


(a) t=2s


(b) t=4s


5 Copyright 2013 by ASME

(c) t=6s

(d) t=8s

(e) t=10s


(f) t=12s

(g) t=14s

(h) t=16s

Figure 7: Snapshots of water volume fraction on the
middle vertical plan at every 2 seconds intervals.
6 Copyright 2013 by ASME
Figure 8 shows a series of snapshots of velocity magnitude
contours on the middle vertical plane. The flow around the
cylindrical structure is captured in the transient CFD
simulation.

(a) t=5s

(b) t=10s


(c) t=15s


(d) t=20s

Figure 8: Snapshots of velocity magnitude on the middle
vertical plane at every 5 seconds intervals.


Figure 9 shows the trajectory of the COG of the cylindrical
structure during the process of going into the splash zone. The
COG Z shows the cylindrical structure goes down. The COG Y
shows the cylindrical structure swings in Y direction, which is
also the wave direction. The COG x shows that there is
essentially no movement in X direction.
7 Copyright 2013 by ASME


Figure 9: Trajectory of the COG of the cylindrical
structure.

Figure 10 shows the linear velocity of the COG of the
cylindrical structure during this process. The velocity in Z
direction reveals that the cylindrical structure has significant
vibration in the splash zone and is moving up and down within
first 10 seconds. The magnitude of the vibration is reduced due
to hydrodynamic damping after 10 seconds. The velocity in Y
direction shows that the cylindrical structure swings back and
forth in the wave direction. The motion of the cylindrical
structure appears to be basically in the plane of Y - Z. There is
little motion in the X direction.

Figure 10: Linear velocities of the COG of the
cylindrical structure.
Figure 11 shows the angular velocity of the COG of the
cylindrical structure during this process. The angular velocity in
Y direction shows that the cylindrical structure rotates
significantly around Y axis due to waves. The rotations around
X and Z axis are shown relatively small.


Figure 11: Angular Velocities of the COG of the
cylindrical structure.

Figure 12 shows the forces on the cylindrical structure
during this process. The crane hook load mainly comes from
the force in Z direction since the forces in X and Y directions
are relatively small compared with the force in the Z direction.
The force in Y direction would represent the slamming force
from the wave. It shows this slamming force changes direction.
The magnitude of the slamming force also changes
significantly. The force in X direction appears very small and
essentially no effect on the cylindrical structure.



Figure 12: Forces on the cylindrical structure.
Figure 13 shows the moments on the cylindrical structure
during this process. The moment around Y axis dominates
while the moments around X and Z axis are very small. This
moment around Y axis causes the cylindrical structure swings
in the wave direction.
8 Copyright 2013 by ASME

Figure 13: Moments on the cylindrical structure.

The numerical results from Figure 9 to Figure 13
quantitatively reveal the behavior of the cylindrical structure
shown in the snap-shot pictures from Figure 7 to Figure 8.
CONCLUSION REMARKS AND FUTURE WORK
This paper demonstrates a solution of two-way coupled
transient CFD and time domain diffraction simulation of a steel
cylindrical structure lowering through the splash zone.
The method of two-way coupled transient CFD and time
domain diffraction modeling of the cylindrical structure
presented in this paper is shown to have more realistic
simulation of a subsea equipment/structure going through the
splash zone. Flow and waves around the structure is simulated
using the VOF model in transient CFD. The hydrodynamic
forces from the flow and waves are being fed into time domain
diffraction model instantaneously, instead of using assumed or
calculated hydrodynamic properties from stead state analyses.
This method eliminates the inaccuracy inherited from
assumed or calculated hydrodynamic properties from stead
state analyses as recommended by DNV-RP-H103. Therefore,
the force in crane wire could be more accurately predicted.
The future work would be an actual offshore installation
case study of a subsea equipment/structure. Comparison of the
result using the two-way coupled method presented in this
paper and the result from traditional time domain diffraction
installation analyses would be provided. Stress and fatigue
damage calculation of the subsea equipment/structure going
through splash zone in different sea states would also be
provided. The recommendations for allowable sea states of
installation would be provided.

ACKNOWLEDGMENTS
The authors would like to thank Deepwater Engineering
Division (Genesis, Houston) and Subsea Structures and
Foundation Department of Genesis, Technip for encouraging
this study and allowing this paper to be published.
REFERENCES
[1] DNV-RP-H103 (April 2011), Modelling and Analysis of
Marine Operations.
[2] Molin B. (2001) "On the added mass and damping of
periodic arrays of fully or partially porous disks". J. Fluids
& Structures, 15, 275-290.


[3] Molin, B. and Nielsen, F.G. (2004) Heave Added Mass
And
Damping of a Perforated Disk below the Free Surface,
19th Int. Workshop on Water Waves and Floating Bodies,
Cortona, Italy, 28-31 March 2004.
[4] Ansys Inc. (2011) ANSYS FLUENT Users Guide,
Release 14.
[5] Ansys Inc. (2011) ANSYS AQWA User Manual,
Release 14.

You might also like