Injection fluid - Compositional Module: GenericOverallCompositionMOdel

[Ras]

Hi Team,

My objective here is trying to understand how injection wells are setup in the compositional module. More specifically, I'm trying to understand the functions that are related to (or use) surface density, comp_i, and components to calculate injection rates. 

I refer to mrst2020b and mainly to the example: compositionalExample3DSixComponents. I run this example once using GenericOverallCompositionModel and once using OverallCompositionCompositionalModel (for the exact same setup). Running these two models, I make the following observations:

• Starting with the generic model, a user input of W.comp_i and W.component forces a flash calculation (prior to simulation) that is based on W.component to calculate corrected/validated comp_i as well as create a new column for rhoS. This is true for injectors and producers. In the case of injectors, these flashed rhoS and comp_i are then used to calculate surface rates.
•  For the overallcomposition model, no flash calculation is called prior to simulation. Therefore, the 'static' surface density from the black-oil properties as well as the comp_i (as was defined by the user) are used to calculate the surface rates.

Based on the above, my questions are as follows:

1. Is the generic model physically more correct to calculate surface rates for injectors as compared to the overallcomposition approach? Is the generic approach analogous to calculating surface rates for producers using the separator group?
2. If from question #1 turns out that the generic model is indeed correct, to have the overallcomposition follow the same approach, I should just flash W.component to surface condition and then ensure b-factor is calculated using this new rhoS?
3. Given that generic model flashes the composition to get comp_i, does this makes the user input of comp_i irrelevant? Unlike in the overallcomposition model where user input of comp_i dictates the composition of the injected fluid.
4. In the generic model, W.component is flashed for all wells regardless of their type (prod or inj). However the resulting rhoS is not relevant in the prod calculation, correct?
5. When running the generic model, W.component goes from representing pure CO2 injection, i.e. [0,1,0,0,0,0] to a zero vector in 'states' for all simulation time steps. This is the result of running initWellSolAD.m prior to simulation. But what does that mean? How does the model identify the composition of the injected fluid?.

Sorry for the lengthy post.

Thanks

[Olav Møyner]

Hi,

1) It depends on what you have set as your rhoS. If you exactly know the rhoS you want to use for injection, you can simply add it to the well as .rhoS and the simulator will use that value. For most users, flashing at the surface to get injection densities will be a reasonable default and this is in line with what most commercial simulators do.

2) Specifying rhoS from flash will give the same result. With the old models you are limited to a single density for all wells, though, so not all cases are possible to simulate with the old model.

3) It is irrelevant if you do not have an aqueous phase just as the initial saturations in the reservoir are irrelevant for the same setup. If a third phase is present, the sum of the two phases governed by the EoS matter.

4) If a well is producing then the surface density in the fluid is used (if no separator is specified). Components are then associated with a phase stream based on their LV-partition as a pure mixture at surface conditions, which is a simplification to avoid re-flashing (but a separator can be used to get a more realistic behavior).

5) The generic models do not output components, but instead output the more general total masses per second in the wellSol which can be used to reconstruct the molar fractions. W.components is still used to set the injection composition.

There's also an example that demonstrates some of this here: https://bitbucket.org/mrst/mrst-autodiff/src/master/compositional/examples/book-ii/compositionalSeparatorExample.m

Best regards,

Olav

https://bitbucket.org/mrst/mrst-autodiff/src/master/compositional/examples/book-ii/compositionalSeparatorExample.m

---

From: sinte...@googlegroups.com <sinte...@googlegroups.com> on behalf of Ras <mra...@usc.edu>
Sent: Sunday, July 4, 2021 21:35
To: MRST-users: The Matlab Reservoir Simulation Toolbox User Group <sinte...@googlegroups.com>
Subject: [MRST Users] Injection fluid - Compositional Module: GenericOverallCompositionMOdel

 

--
You received this message because you are subscribed to the Google Groups "MRST-users: The Matlab Reservoir Simulation Toolbox User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sintef-mrst...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sintef-mrst/0b249494-1239-4585-ac75-62463b0b8e57n%40googlegroups.com.