3D CFD Simulation of Flare Flow and Its Thermal Impact Analysis On The Stopple For Flare Tip Replacement

Discipline BP BERAU LTD.
Document Number
Process & Process Safety TGH-086-REG-PS-BEI-0001
Project/PO No.
4800082713
Author’s Org.
Berca Engineering International Next Revision Date
NA
3D CFD Simulation of Flare Flow and Its Thermal Impact

Analysis on the Stopple for Flare Tip Replacement
Assumption Register
Company Approval
Name Title Date Signature
Technical Approval
Rev Date Issue Description Prepared By
Approval Authority
A 22-07-21 Issued for Review LMS TC TC
3D CFD Simulation of Flare Flow and Its Thermal Impact Analysis on the Stopple for Flare Tip Replacement–
Assumption Register
DISCLAIMER
Gexcon shall not be liable for damages, which the assignor, or assignor’s clients, vendors,
consultants or other third party, suffers when applying or using the results of Gexcon’s work,
unless there is misconduct or gross negligence on the part of Gexcon or on the part of the persons
used by Gexcon to carry out the work.
TGH-086-REG-PS-BEI-0001
Assumption Register
TABLE OF CONTENT
DISCLAIMER..................................................................................................................................2
TABLE OF CONTENT...................................................................................................................3
1.0 GENERAL ASSUMPTION.................................................................................................4
2.0 GRID GENERATION AND SETUP..................................................................................6
3.0 REFERENCE......................................................................................................................14
APPENDIX A – Geometry Model (Additional Figures).............................................................15
Assumption Register
1.0 GENERAL ASSUMPTION
Item General Assumption

1.1 Geometry Model
The 3D geometry model of Tangguh Dry Flare was provided in Naviswork format [1]. Figure
1 -1 displays top view of pipeline from train-1 to KO Drum. Complete dimensions of
geometry in Figure 1 -1 can be seen in Appendix A. The technical drawing in Appendix A
was provided in [2]. Figure 1 -2 displays side view, back view, and dimension of KO Drum.
Complete dimensions of KO Drum was provided in [3,4,5]. The geometry information
obtained from both references is used to re-model a geometry from 20m upstream and 20m
downstream T junction to stopple location in a format that is compatible with ANSYS
(STEP).
Stopple
Figure 1-1 Top View of Pipeline from 20m upstream and downstream T junction to stopple.
Figure 1-2 Side View, Back View, and Dimension of KO Drum
The cryogenic fluid is a multiphase fluid. To achieve highest accuracy of multiphase CFD
results, the CFD is modelled in 3D domain without any symmetry boundary condition or
geometry simplification. The actual geometry provided is on the left side of Figure 1 -4 and
Figure 1 -4. The re-modelled geometry is on the right side of Figure 1 -4 and Figure 1 -1
Figure 1-3 Actual geometry and re-modelled geometry
Assumption Register
Item General Assumption
Figure 1-4 KO Drum actual geometry and re-modelled geometry
The internal arrangement of KO Drum will not be modelled in detail. In Figure 1 -5, a circle
face is a representation of injection gas inlet. This face located near the stopple as its function
to maintain the temperature of the stopple.
Injection gas inlet
Figure 1-5 Injection Gas Inlet Geometry Representation: Isometric View

1.2 Other Assumptions
The electric heater inside the KO Drum will be deactivated during process up set and
therefore will not be considered in the modelling.
Assumption Register
2.0 GRID GENERATION AND SETUP
Item Grid Generation and Setup

2.1 Grid Generation
Grid generation is the process of discretizing the fluid/solid domain into small domains. The
small domains are control volume used in the derivations of fluid dynamics and heat transfer
equations. There are two types of grids, namely structured and unstructured grid.
Structured grid is a grid whose elements and nodes are easy to be identified. Structured grid
usually has orthogonal elements which is the most accurate element in CFD or FEA.
Structured grid is memory efficient and fast to be solved but hard to be generated on complex
geometries. Angled and curved geometries are also approximated in structured grid (cannot
have fully orthogonal elements).
Unstructured grids have arbitrary elements connectivity. The connectivity needs to be stored
and therefore needs greater memory. Despite that fact, unstructured grid is easier to be
generated on complex geometries.
Figure 2-6 Example of Tetrahedral Grid
In ANSYS, both the generation of structured and unstructured grid can be done. Unstructured
grid is selected to be used because it requires less handling time. There are several types of
element in unstructured grid, such as tetrahedral, hexahedral, and polyhedral. Tetrahedral
element fits curved geometry very well than hexahedral element.
Polyhedral element which is generated from tetrahedral element is selected to be used in the
simulation. Polyhedral grids have higher accuracy in recirculating flows and lesser cell counts
than tetrahedral grids which can reduce computational time. An example of tetrahedral grid is
displayed in Figure 2 -6 and an example of polyhedral grid is displayed in Figure 2 -7.
Assumption Register
Figure 2-7 Example of Polyhedral Grid
In a fluid flow, the velocity of the fluid at freestream area and at area near a wall (surface) will
be different. Near the wall, the velocity decreases non-linearly up to a point where the fluid
will have zero velocity at the wall which is caused by shear stress. This is what is termed the
“no slip” wall condition in CFD.
If it is important to report separation points or pressure drops accurately, then the gradient
near the wall needs to be captured correctly. To do this, inflation mesh or empirical wall
function can be used. Inflation mesh has more accurate resolution of the boundary layer (the
gradient layer) than the empirical wall function but has more computational expense since it
increases number of mesh cells.
In this project, separation points or pressure drops are not being concerned. Hence, empirical
wall function is chosen to capture the gradient. By doing so, more computational expense can
be avoided. Representation of Wall Function approach vs fully resolving the boundary layer is
displayed in Figure 2 -8.
Figure 2-8 Representation of Wall Function Approach vs Fully Resolving The Boundary Layer
[7]
2.2 Material
Cryogenic fluid is modelled as a mixture material because it consists of several chemical

species. The use of mixture material will determine several physical properties of the mixture,
such as density, viscosity, and thermal conductivity which are functions of the composition. It
is also modelled as multiphase since it consists of vapor and liquid (solid is assumed
negligible if there is any). It is assumed that there is no reaction in the mixing problem but
there can exist phase changes (evaporation or condensation) influenced by pressure and
Assumption Register
temperature in the flow field.
Material choices of gas injection are N 2 and propane. Material with higher mass flow rate to
maintain temperature near stopple will be selected to be used in the simulation. Based on
current assessment, propane is proposed to be used in the simulation due to the following:
 Propane has higher mass flow rate than N 2 (with the same pressure and temperature)
because propane molecular weight is bigger than N2 molecular weight.
 Propane causes more temperature decrease than N 2 because propane heat capacity is
higher than N2 heat capacity.
Mass flow rate per area difference between propane and N 2 at a same operating condition is
67.66% with respect to N2 mass flow rate per area as can be seen in Table 2 -1.
Table 2-1 Propane and N2 Mass Flow Rate Per Area Comparison
Gas Propane N2
Operating Pressure (Bara) 6.1 6.1
Operating Temperature, Celcius 80 80
Molecular Weight, g/mol 44.0956 28.0134
Vapour Density, kmol/m3 0.220735 0.207235
Density, kg/m3
9.733442 5.805357
(Molecular Weight × Vapor Density)
Exit Velocity, m/s 18 18
Mass Flow rate per area, kg/m2/s 175.202 104.496
Mass Flow rate per area difference percentage with
67.66
respect to N2, %
*
Source: Phast
2.3 Cases and Boundary Conditions
To analyse the thermal effect of possible cryogenic exposure to the stopple, liquid profile from
KO Drum to stopple, as well as the effectiveness of gas injection in maintaining temperature
near the stopple within allowable limit, a total of 6 cases will be simulated in transient time, 3
consecutive cases without gas injection and 3 consecutive cases with gas injection. Cases
summary is displayed in Table 2 -2. In each case, there will be four boundary conditions,
namely two inlets, an outlet, and a wall. T
Table 2-2 Cases Summary

Case Operating conditions Initial condition Duration
1  Train-1 and Train-2 Power Failure Purging 3 hours
 Without gas injection
2  Train-1 Start-up with Train-3 Normal Operation From Case 1 13 hours
 Without gas injection except for
pressure
3  Train-1 Start-up Together with Train-3 Power From Case 2 5 hours
Failure
 Without gas injection
4  Train-1 and Train-2 Power Failure Purging 3 hours
 With gas injection
5  Train-1 Start-up with Train-3 Normal Operation From Case 4 13 hours
 With gas injection except for
pressure
6  Train-1 Start-up Together with Train-3 Power From Case 5 5 hours
Failure
Assumption Register
 With gas injection
The first inlet is located at 20m upstream the Train-1 T junction. The second inlet is located at
gas injection orifice. For cases without gas injection, this inlet will be treated as a wall (no slip
condition). The wall boundary represents the pipeline wall (no-slip condition) and stopple
surface (no-slip condition). The wall boundary is assumed as an adiabatic wall (no heat
transfer). The outlet is located at 20m downstream the Train-1 T junction. The 20m distance is
chosen as a distance that is far enough from the first obstacle from inlet/outlet (a pipe bend) so
that the flow recirculation is not captured at the inlet/outlet. Flow recirculation can result in
longer time of reaching convergence in the simulation. The sketch of computational domain
and its boundary conditions is displayed in Figure 2 -9.
Outlet
Inlet Gas injection inlet
Figure 2-9 Sketch of Computational Domain and Its Boundary Conditions
The mass flow-inlet and pressure-outlet type will be chosen for the inlet and outlet boundary
because it is suitable for the input data which will be provided. This pair of inlet and outlet
type is also robust. The pressure-inlet type is chosen for the inlet boundary at injection gas
orifice because it is suitable for the input data which will be further assumed, that is pressure
and temperature. Case 1 until 6 is explained in detail in Table 2 -3.
Table 2-3 Case 1-6

Case 1-3 (Without injection gas)
Case 1 Train-1/2 Power Failure – Without Injection
The temperature distribution near the stopple under Train-1 and Train-2 power failure without gas
injection will be analyzed within 3 hours duration. The initial condition (yellow area) is based on
purging condition.
Case 2 Train-1 Start Up & Train 3 Normal Operation – Without Injection
Assumption Register
The temperature distribution near the stopple under Train-1 start up with Train-3 normal operation
without gas injection will be analysed within 13 hours duration. The initial profile is based on
modelling profile results of Case 1 except for pressure that is equal to 0 barg.
Case 3 Train 1 Start Up Together with Train-3 Power Failure – Without Injection
The temperature distribution near the stopple under Train-1 start up together with Train-3 power
failure without gas injection will be analysed within 5 hours duration. The initial profile from T
junction to stopple (purple area) is based on modelling profile results of case 2.
Case 4-6 (With injection gas)
Case 4 Train-1/2 Power Failure – With Injection
The temperature distribution near the stopple under Train-1 and Train-2 power failure condition
with gas injection will be analysed within 3 hours duration. Required gas injection mass flow rate
should be determined through iterative simulation. The purpose of gas injection is to maintain
temperature near stopple within the allowable limit. The gas will be injected when the temperature
indicated by TI (Temperature Indicator) reaches low temperature set point and stopped when the
stopple temperature reaches high temperature set point.
Case 5 Train-1 Start Up & Train 3 Normal Operation – With Injection
Assumption Register
The temperature distribution near the stopple under Train-1 start up with Train-3 normal operation
with gas injection will be analysed within 13 hours duration. Required gas injection mass flow rate
should be determined through iterative simulation. The gas injection will be injected when the
temperature indicated by TI reaches low temperature set point and stopped when the stopple
temperature reaches high temperature set point.
Case 6 Train 1 Start Up Together with Train-3 Power Failure – With Injection
The temperature distribution near the stopple under Train-1 start up together with Train-3 power
failure with gas injection will be analysed. The duration is 5 hours, and the initial condition is based
on results of case 5. The gas will be injected when the temperature indicated by TI reaches low
temperature set point and stopped when the stopple temperature reaches high temperature set point.
2.4 Output Summary
Table 2 -4 summarizes outputs that will be obtained in this project.
Table 2-4 Output Summary

Type of Output Obtained from Remarks
Temperature profile near stopple for all cases CFD Result
Liquid profile from KO Drum area until stopple area for CFD Result
all cases
Gas injection for Case 4-6
 Material Properties study
 Time CFD Result
 Duration CFD Result
 Mass flow rate CFD Result
 Frequency CFD Result
2.5 Input Summary
Table 2 -5 summarizes inputs that will be used in the CFD simulation.
Table 2-5 Input Summary
Assumption Register
Type of input Provided by Default values Remarks

BP/BEI to be used
Geometry
 3D geometry Complete
 Isometric drawing from stopple to T Complete
junction
 Pipeline dimensions upstream T Complete
junction (isometric drawing)
 Location of injection gas Complete 500mm
 Diameter of injection gas inlet Complete Based on
velocity of 18
m/s.
Case 1
 Cryogenic mass fraction/mole fraction Complete
at inlet
 Properties1 of each cryogenic specie at Complete
inlet
 Liquid and vapor variables2 at inlet Incomplete

 Pressure at inlet Complete

at outlet
outlet
 Pressure at outlet Complete

Incomplete
 Liquid and vapor variables2 at outlet
Complete
 Cryogenic mass fraction/mole fraction
at initial time
Complete
 Properties1 of each cryogenic specie at
initial time
Complete
 Pressure at initial time Complete
 Temperature at initial time Complete
 Velocity at initial time
Case 2
at inlet
inlet
 Liquid and vapor variables2 at inlet Incomplete

 Pressure at inlet
Incomplete Initial is not
 Cryogenic mass fraction/mole fraction complete
at outlet Complete
outlet Complete
 Pressure at outlet
Complete
 Liquid and vapor variables2 at outlet
Incomplete
Case 3
Assumption Register

at inlet
inlet
 Liquid and vapor variables2 at inlet Incomplete Initial is not

complete
 Pressure at inlet
Incomplete Initial is not
 Cryogenic mass fraction/mole fraction complete
at outlet Complete
outlet Complete
 Pressure at outlet
Complete
 Variables2 at outlet
Complete
Injection gas 0.1 bar above
 Pressure Complete line pressure
80 0C
 Operating temperature Complete 18 m/s
 Maximum velocity Complete N2, Propane
 Material choices Complete
Temperature indicator
 Low temperature set point Complete 5 0C
 High temperature set point Complete 10 0C
 Duration injection from low Complete 10 minutes
temperature set point to high
temperature set point
 Location of temperature indicator Complete 13 m from
Stopple
1
Chemical formula, specific heat, thermal conductivity, viscosity, molecular weight, density, saturation
pressure, saturation temperature
2
Mass flow, temperature, velocity, specific heat
3.0 REFERENCE
1. 3D Model (Navisworks format), BP_Tangguh_all_05-11-2020.nwd, received on 5 Nov

2020.
2. Technical Drawing (pdf format), Isometric dry.pdf, received on 9 Jun 2021.
3. Technical Drawing (pdf format), 086-VDR-BBI-1207_R3B_GAD Dry KO Drum, received
on 9 Jun 2021.
4. Technical Drawing (pdf format), 086-VDR-BBI-1252_R3A_Nozzle Detail Dry KO Drum,
received on 9 Jun 2021.
5. Technical Drawing (pdf format), 86-EDS-VM-1261_R6B_Eqp Datasheet Dry KO Drum,
received on 9 Jun 2021.
6. ANSYS FLUENT 12.0/12.1 Documentation
7. “TIPS & TRICKS: INFLATION LAYER MESHING IN ANSYS”,
https://www.computationalfluiddynamics.com.au/tips-tricks-inflation-layer-meshing-in-
ansys/, accessed on 8 July 2021.
Assumption Register
APPENDIX A – Geometry Model (Additional Figures)
Figure A-1 D250-086-VD-1001-34”-1SA0K-NI (technical drawing) -- isometric
Assumption Register
Assumption Register

3D CFD Simulation of Flare Flow and Its Thermal Impact Analysis On The Stopple For Flare Tip Replacement

Uploaded by

Copyright:

Available Formats

3D CFD Simulation of Flare Flow and Its Thermal Impact Analysis On The Stopple For Flare Tip Replacement

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3D CFD Simulation of Flare Flow and Its Thermal Impact Analysis On The Stopple For Flare Tip Replacement

Uploaded by

Copyright:

Available Formats

Discipline BP BERAU LTD.

3D CFD Simulation of Flare Flow and Its Thermal Impact

1.0 GENERAL ASSUMPTION

Item General Assumption

Figure 1-2 Side View, Back View, and Dimension of KO Drum

Figure 1-3 Actual geometry and re-modelled geometry

Item General Assumption

Figure 1-4 KO Drum actual geometry and re-modelled geometry

Injection gas inlet

Figure 1-5 Injection Gas Inlet Geometry Representation: Isometric View

2.0 GRID GENERATION AND SETUP

Item Grid Generation and Setup

Figure 2-6 Example of Tetrahedral Grid

Figure 2-7 Example of Polyhedral Grid

Cryogenic fluid is modelled as a mixture material because it consists of several chemical

temperature in the flow field.

2.3 Cases and Boundary Conditions

Table 2-2 Cases Summary

 With gas injection

Inlet Gas injection inlet

Figure 2-9 Sketch of Computational Domain and Its Boundary Conditions

Table 2-3 Case 1-6

2.4 Output Summary

Table 2 -4 summarizes outputs that will be obtained in this project.

Table 2-4 Output Summary

2.5 Input Summary

Table 2 -5 summarizes inputs that will be used in the CFD simulation.

Table 2-5 Input Summary

Type of input Provided by Default values Remarks

 Liquid and vapor variables2 at inlet Incomplete

 Cryogenic mass fraction/mole fraction Complete

 Pressure at outlet Complete

 Liquid and vapor variables2 at inlet Incomplete

 Cryogenic mass fraction/mole fraction Complete

 Liquid and vapor variables2 at inlet Incomplete Initial is not

1. 3D Model (Navisworks format), BP_Tangguh_all_05-11-2020.nwd, received on 5 Nov

APPENDIX A – Geometry Model (Additional Figures)

Figure A-1 D250-086-VD-1001-34”-1SA0K-NI (technical drawing) -- isometric

Figure A-2 D250-090-VD-1001-32”-1SA0K-NI (technical drawing) -- isometric

Figure A-3 D260-090-VD-1001-32”-1SA0K-NI (technical drawing) -- isometric

You might also like