¦ SIMULATIONS ¦
Plan of this page :
1. Vertical pipe
2. Network of pipes
3. Comparison Tacite / Olga, nb 1
4. Comparison Tacite / Olga, nb 2
5. Comparison Tacite / Olga, nb 3
This simulation was the first one I implemented with Olga.
I had previously performed a similar simulation of two phases flows under a vertical pipe with TACITE
and I decided to try to perform the same simulation with Olga 2000.
The purpose of both simulations was to establish configuration card of the flows function of the surface velocity of,
respectively, gas (jg) and liquid (jl). But the results are not presented because the tests were unphysical and probably
not totally similar
du to problems in the different PVT files. So only the way this test can be implement is presented there.
I studied a vertical pipe of 100 meters length.
At the entry, I implemented a source at 24°C and I changed the mass flow and the fraction of gas to obtain
the configuration card.
At the exit, I introduced a outlet at 24°C (no heat exchange) and 40 bar.
To have 24°C at inlet and outlet, I stated 'temperature = off' in the operating
condition and I create the 'initial condition' module with 24°C as temperature at the beginning.
ClickHERE to see the .inp file
Normally, configuration cards are draw as function of jg and jl, respectively the surface velocity of gas and liquid.
These surface velocities can be obtained with the following formula:
For Olga, at the entry, the total mass flow rate (Qt) and the gas fraction (Gfraction) have to be given :
PS : For TACITE, users have to give Qg and Ql the mass flow rate of gas and liquid at the entry.
With this test, I encountered the first real difficulties since this case is not presented in the Olga manual.
Besides, I also have to implement the PVT file.
It appears that problems still remind in this case so the comparison with Tacite was aborted for this case.
Network of pipes
This simulation had been performed for the BEI ( Industrial Project Simulation) 'Two phase flows and petroleum transport' by Julian Calvo, Amelie Baudon and I.
It consisted in the implementation of a network of pipes lying four wells to a manifold which collects the flows and sends they to a FPSO with one pipe presenting a 90° angle.
Only the creation of the geometry is presented here since the BEI will have its own web site.
We considered a network of 5 branches: the four first link a well to the manifold and the fifth links the manifold to a FPSO.
Branches one to four are similar: they are constituted by two pipes of 5+3/8 inches. The first is vertical and 800 meters long.
It is between the ground of the sea and the well. The second pipe is horizontal lying in the ground of the sea. It is 100 meters long and its extremity is the manifold.
PS: To create the four branches, we duplicated the branch one and only change the starting point of the branch.
The fifth branch is composed by two pipes of 8 inches. The first pipe is horizontal (lying in the ground) and its length is equal to 10 km.
The second pipe is vertical and its length is equal to 1 km. Its extremity is the FPSO.
The following graphs have been obtained via drawing the profile of branch 2 and 5 under Olga 2000.
To do that, we had only press on the branches, press the right button of the mousse and select 'draw profile'. These graphs show the geometry of the divers branches.
The total geometry is represented by Olga 2000 via horizontal arrows merging in the manifold defined as a merge node while the entree_1 to entree_4
and the sortie are defined as terminal nodes .
The following graph is a screen capture of the Olga case background. It is the representation of the network by Olga.
Note that Olga draws always branches with blue horizontal arrows (it was also the case for the vertical pipe). Consequently, to verify the geometry of branches,
the draw profile
option is useful (click on the branch to inspect, click on the right button of your mousse and select 'draw profile').
To see the .inp file, click HERE
Comparison Tacite / Olga, nb 1
The vertical case presented in the first simulation had been reused with an other PVT file.
This case is a comparison between Olga and Tacite for a vertical pipe of 100 meters with a water and methane flow.
The final pressure is set to 40 bars and the temperature is set to 23°C (no thermal exchange).
This case had been chosen because it is quite well known. Configuration cards had been drawn for jg between 0.2 and 10 and jl between 0.2 and 4.
Then, a local study had been performed for jg = 4 m/s.
ClickHERE to see the .inp file
ClickHERE to see the .tab file
The configuration cards are very different : Olga predicts essentially bubble flow regime and Tacite predicts
essentially slug flow regime. At jg = 4 m/s, for Olga, the flow regime transition is around jl = 1m/s although,
for Tacite, the flow regime transition is around 3.5. In fact, experiments are between these two results. It appears that Tacite transition is a bite high but the Olga one is a too low
( in particular, for high gas flux, Olga transition is totally not in the good range).
The other results are however very similar.
The following graphs present the results of Tacite and Olga function of jl when jg had been set to 4m/s.
Comparison Tacite / Olga, nb 2
This case is also a vertical pipe of 100 meters. The PVT file is more complex: one gas and 3 liquids.
This PVT file is the one that had been used in the industrial case simulation project (called BEI and presented in
the second simulation : 'network of pipes'). This simulation did not introduced thermal exchange: temperature had been set
to 23°C. The .inp file is similar to the previous case.
The following graphs present the variation of the pressure loss and the variation of exit void rate in function of jl.
The results are very similar even if the flow regime transition is around 0.8 for Olga and 1.8 for Tacite.
Comparison Tacite / Olga, nb 3
This case is similar to the previous one: same geometry (vertical pipe of 100 meters) and same fluids.
In this case, thermal exchanges had been implemented. Three similar cases are presented below :
- 1st case: external temperature = 5°C, mass flow rate = 10 000 barrels/day (normal conditions),
- 2nd case: external temperature = 5°C, mass flow rate = 20 000 barrels/day (normal conditions),
- 3rd case: external temperature between 5 and 15°C, mass flow rate = 20 000 barrels/day (normal conditions).
ClickHERE to see the .inp file
The following graphs are drawn for normal condition (stoechiometric rates) in function of the position in the pipe.
Below, the cases are represented by colors:
- For Olga the first case presents a bubble flow although Tacite computes a slug flow. For the two other cases, Tacite and Olga computes bubble flows.
- The results are very similar and the cases 2 and 3 often gave the same results (curves are superposed).
Lastly, a simulation similar to the 3rd case was performed. The differences were the following:
thaks to this simulation, the temperature along the pipe really deacreases. The results of Tacite and Olga in mater of temperature are:
- Wall thikness = 0.000001 m (and not 0.049),
- Hambient = 200 W/m/m/C.
We can observe that the difference between the two hardwares is only around 1°C (but it is also 10% !).