T2Well-ECO2N
Dear T2Well users; Looking for a tool to design CO2 injection aiming at CCS. T2Well-ECO2N software is new to me, I wonder if it is only a 1D simulator? Do you have good experience with modelling near well effect of CO2 injection (for example in depleted reservoirs or else)? Is there any existing code coupling a T2Well module with Tough2 where the entire suite of EOS including EOS2 or EOS7C functioning in a 3D setting? I very much looking forward to and appreciate your possible response.
Thank you.
22 replies
-
Dear Godarzi,
T2Well-ECO2N allows simulating the coupled wellbore-reservoir flow. Of course, while 1D flow is modeled along the wellbore, the reservoir can be discretised as usually made for standard TOUGH2 simulations, including 3D cartesian or unstructured grids.
T2Well-ECO2N is perfectly suitable for modeling near wellbore processes in full transient conditions.
When you mention depleted reservoirs what are you referring to? You cannot model CO2 injection in depleted natural gas reservoirs with T2Well-ECO2N, as it handles only H2O-NaCl-CO2 mixtures.
As far as I know T2Well was already coupled at LBNL also to ECO2M, EOS3, and an enhanced version of EOS8. I have worked with T2Well coupled to EWASG and EOS2H, the latter a version of EOS2 which is able to model steam-like supercritical conditions.
Coupling T2Well to one of the standard TOUGH2 EOS modules, such as EOS7C, is rather simple once you have access to both source codes.
Regards,
Alfredo
-
Thank you Alfredo,
Your words made me to dig up a bit more. But still, I am in search for a non-isothermal Tough-based coupled well & reservoir modelling solution to model brine (or water+ NaCl)+Co2(all phases)+HC gases (at least CH4) including partitioning of CO2 in both gaseous and aqueous phases together with salt precipitation. I would rather avoid ToughReact if possible. I appreciate your further comment on this topic. A summary of what I could capture in Tough documentations regarding this topic follows:
EOS7C handles Brine, CH4-CO2 mixtures, CO2 in only non-condensable form, no phase transitions and no solid precipitation, 2 phases (G+Aq) and 5 components (W+Br+NCG+Tracer+CH4)
ECO2N handles H2O-NaCl-CO2 mixtures, CO2 in only non-condensable form, no phase transitions but solid appearance/disappearance, no HC gases, 3 phases (G+Aq+S) and 3 components (W+NaCl+CO2), P:1000 bar, T: 10-300’C
ECO2M handles H2O-NaCl-CO2 mixtures, CO2 in both gas and liquid phases, phase transitions and solid appearance/disappearance, no HC gases, 3 phases (G+Aq+S) and 3 components (W+NaCl+CO2), P: 600bar, T:10-110’C
I am aiming at using depleted gas field for CCS application. Is it right to say that a suitable EOS could be a step away from ECO2M due to the lack of CH4? I am not sure if there any ongoing EOS improvement to look for in the near future or it might be better idea to look for a Tough based simulation with something else; like Real Gas Brine (Moridis / Freeman??)?. Also important to see whether T2Well couplable to ECO2M as easy as ECO2N?
Best regards
Godarzi
-
Dear Godarzi,
To help you decide whether you need ECO2M versus ECO2N, below is the phase diagram of CO2. If your problem will not cross the saturation line (shown in black) or get near the critical point (the end of the saturation line at about 31 C and 74 Bar), then ECO2N will work for you. But it does not have CH4, so that may be a deal breaker. If you opt for EOS7C, then you can have CH4, but no salt precipitation, and you still need to not cross the saturation line or get close to the critical point. The only EOS coupled to T2Well that is available to the public is ECO2N, but Alfredo may be able to point you to the people who have created other versions for their own use on particular projects, and they might be willing to share them with you. I am not so familiar with Real Gas Brine, but according to the user's guide, you can have CO2, CH4, and salt that can precipitate. CO2 is described as a "condensable" gas, but there does not appear to be a provision for more than two fluid phases (aqueous and gas), so it may be that you cannot cross the saturation line there either. I don't see any mention of a wellbore simulator in Real Gas Brine, but it does include an option for non-Darcian turbulent flow, which you might find reasonable to represent wellbore flow.
Christine
-
Thank you Christine and Alfredo;
for useful advises and pointing out rational behind the EOS’s constraints. I hope that I will be able to come in contact with people aiming at similar applications, including the ones that might have coupled ECO2M.
However, With cold- but deep injection and heterogenous anticline structure, I expect entering the P/T envelope around critical point is inevitable. We have also to include a supercritical phase, whether dealt with as a “dense unique phase” or a liquid within an extended physio/thermodynamic property envelope. I have to peruse further what you indicated on RGB, regarding handling of CO2 phases. From the manual you get the impression that CO2 partitioning finds place properly??
The other thing that I try to find out is the scalability and simulation efficiency when you undertake a storage project of huge dimension that should be modelled and run for at least 10 decades. Here, it might come the question of using Though2, Tough+ or other alternative able to use effectively 10s/100s of cpu cores in parallel. I appreciate any reference or advice on that front as well.
Best regards
Godarzi
-
ECO2M has been already coupled to a new, 3-phase flow version of T2well coded by Lehua Pan. It can handle all the phase combinations of H2O-NaCl-CO2 mixtures, including halite precipitation. But it cannot handle CH4.
If you really need 3-phase flow with multicomponent inorganic and organic fluid mixtures, I think the only code of TOUGH family able to do that is TOGA by Lehua Pan. Its distribution by LBNL started in 2020.
But TOGA should be coupled to the new 3-phase version of T2Well for coupled wellbore-reservoir flow simulations.
T2Well implements a full transient approach for wellbore flow which is in particular suitable for simulating the wellbore transients at injection startup or shut-in. For long term field-wide simulations of CCS with multiple wells, T2Well might increase the simulation time because of tiny time steps required for solving strong wellbore transients.
Years ago I coded wellbore flow equations in a TOUGH2 version assuming wellbore flow was steady-state within each time step. The solution was not so accurare as that of T2Well for injection start-up, but was satisfactory when modeling longterm injection with slow transients.
That version is coupled to the TMGAS EOS which can handle mixtures of inorganic gases and HCs (C1-C10, BTEX, +pseudo-components for C10+).
The non aqueous mixture can be in gas, liquid or supercritical conditions, but only two-phase brine - non-aqueos phase equilbrium can be simulated. That is only one NAq phase can be present at the same time in one grid element.
We run injection of CO2-rich mixtures in gas reservoirs and saline aquifers and of sour or acid gas mixtures in undersaturated oil reservoirs (see Battistelli et al., 2010; Castelletto et al. 2013 IJGGC; Battistelli et al., 2011 TiPM).
This year I'll probably work on the coupling of TMGAS to T2Well. A 3-phase version of TMGAS was always considered, but never developed...
-
Following your recent discussions, I have some questions regarding the use of T2WELL . Specifically, if our focus is solely on long-term simulation post-injection, without considering the transient phenomena near the wellbore, would ECO2N or ECO2M be sufficient for our needs? In this scenario, is the use of T2WELL necessary, or can we proceed without it?
Thank you for your guidance and assistance.
Best regards
-
You should use T2Well when a coupled modeling of wellbore and reservoir flow is necessary. This might be the case when you are interested in transients occurring because of changes in injection parameters, such as injection rate and/or enthalpy. Wellbore flow may also be important to:
i. inject CO2 into the reservoir at the proper BHT, which depends on wellbore flow processes too;
ii. to have the proper flowing P gradient inside the wellbore which is for sure different (lower) than the P gradient inside the aquifer, resulting in more CO2 injected at the base of the aquifer with respect to the top.
For i) if you can have a reasonable estimate of BHT of injected CO2, you may inject directly into the reservoir at that T, without modeling wellbore flow. Take into consideration that the increase of BHP would imply a pressurization inside the wellbore and likely a change in T over time due to JT effect.
For ii) if you have supercritical CO2 flow (as you should have by injecting in a sufficiently deep aquifer), single-phase wellbore flow can be conveniently simulated by using Darcy's law and an equivalent permeability of the column of elements that should be used to simulate the wellbore. The equivalent permeability can be estimated using the equation below (Battistelli et al., 2011, TIPM).
You mention your purpose is long-term simulation of post-injection. I guess in this case there is no need to use a coupled wellbore-reservoir approach, providing that reservoir processes during the injection period have been simulated properly and you have a good spatial distribution of P, T, saturations, concentrations of CO2 and salt, to be used as initial conditions.
Regards,
Alfredo
-
Thank you very much for your detailed explanation!
-
-
By the way, I do not think TOUGH+RGB can handle 3-phase flow.
-
TOGA does handle three-phase flow, and when I asked Lehua about the possibility of using it for CO2 in liquid, gaseous, and supercritical phases (like ECO2M), he said that formally it should work, but that he has not tested it.
Regarding your other question about running in parallel, TOUGH3 and TOUGH+ can be run in parallel, but the current version of TOGA is not designed to do that.
-
Thank you both for important additional information.
I feel that I have now some clue about the capabilities of the codes. Your input have been very helpful. I will definitely have to learn more and follow further development of, not least- your well flow module. It is particularly interesting at the stage when we will estimate large dimension storage and need speedy cpu with less transient around the sinks/sources. My prime aim now is to be able to model the near wellbore phenomenon when dry CO2 is injected into very saline aquifer or deep depleted HC formation. This will also mean estimating water vaporization and resulting salt precipitation (reduced injectivity) under transient condition and frequent shut-in events, which is expected in such projects.
Thanks again to both of you.
Godarzi
-
Dear Godarzi,
if you are interested primarily to simulate fast transients at well start-up or shut-in, then you definetely have to use T2Well.
For pure CO2 injection in saline aquifers with T<100°C and single fase flow within the wellbore, T2Well-ECO2N, already distributed by LBNL, can do the job.
T2Well-ECO2M can also simulate 3-phase flow (including the wellbore) and T up to 300°C. I think it is still a research code not yet distributed by LBNL. A search on the TOUGH web site can confirm that.
We are working with T2Well-ECO2M evaluating the code capabilities for CO2 sequestration in saline aquifers. Some results with a slightly modified code version are documented by Strpić et al. (2020). "Modeling of transient multiphase flow in a CO2 injection well with the wellbore-reservoir coupled simulator T2Well-ECO2M". GET2020, 1st Geoscience & Engineering in Energy Transition Conference.
As previously mentioned, when other components are involved (such as other inorganic gases or hydrocarbons), no rediliy available tools seems to be available within the official LBNL TOUGH2/3 family of codes.
I suggest to have a look to Shabani and Vilcaez (2018) for an EOS module for mixtures of CO2-CH4-H2S-N2 (Comp. & Geosc.) and to Loeve et al. (2014), GHGT-12, for an EOS module which includes the handling of CO2-CH4 3-phase mixtures.
Regards,
Alfredo
-
Very good summary. Thank you for the references.
I hope that I get the privilege to contact you in case I need a specific additional information. With what I have got so far, I am really thankful.
Best regards
Godarzi
-
Back with more questions :-)
Thought to ask you about inclusion of "Joule-Thomson" effect in T2Well both over chokes (or wellbore restrictions) and porous plug (sand face). Is this thermal phenomenon either standalone or in connection / association with non-darcy condition in highly heterogenous completion zones have been addressed by any version of "Tough"?
Cheers
Keshvad
-
To extend the question,
If temperature goes below 10'C or in negative territories as a result of J-T effect, how ECO2N / ECO2M handle it then?
Thanks in advance
Godarzi
-
1) JT effects: they are implicitely modeled by ECO2M and ECO2N thanks to the dependency of enthalpy on P&T. So, a depressurization of CO2-rich phase will involve a cooling, while an increase of P will generate heating. This at least at the customary P&T conditions of interest for a common CO2 sequestration project. At increasing P the JT coefficient of CO2 declines, to become negative at very high P (at about 660 bar at T=100°C).
I'm not sure about the code options to model chokes or other flow restrictions. Something I still have to look at.
2) cooling below the freezing T of water.
ECO2M and ECO2n have no provisions to model water freezing, nor the capability to model CO2 hydrates.
In ECO2M the properties of CO2 phases can be modeled accurately down to -20°C. If the aqueous phase is present, its properties and the mutual CO2-H2O solubilities are computed at the fixed T of 3°C when the local T declines below that threshold.
If negative T occurs within the injection wellbore only (not in the reservoir), the approximation seems to be acceptable as customarily no aqueous phase should be present in the injected stream.
Regards,
Alfredo
-
Excellent, a good assurance, thank you for clarifications. Good to hear that it still cover the entire realistic operational envelope of P/T and it has a boundary definition (3'C)-, perhaps good to avoid numerical problems.
I have seen Y. Zhang's "What Could We See at The Production Well Before The Thermal Breakthrough?", an exercise looking into the other end of JT effect, increase of T in a producer as a result of a steady state CO2 injection. Is there anywhere (manual, paper, etc) where the near well cooling (as a result of JT effect) plus possible outcome of thermal fracturing has been observed or brought up?
A fair assumption would be that the T2well-ECO2M approach inside well will follow the same principal relation between P/T and Enthalpy. Would be nice to see if the options to include chokes or restriction like gravel pack are there.
Thank you again Alfredo for your prompt reply and extended follow up. I am quite sure you have many things to use your time on.
Best regards
Godarzi
-
Dear Alfredo
I did a revisit of your earlier comments, since I am looking forward to see the development of T2Well-ECO2M. I still can’t see it in the “lbl.gov” pages. Your publication (Kristina & Battistelli) together with few others are filling an existing gap in research that is significant. Hopefully this is coded in properly, the sooner the better. I wanted just to say hello and hear perhaps from you on the status of code development.
Godarzi
-
Thank you for the update, Alfredo.
Good luck with all the implementation and improvements you are doing at U. of Bologna. I am sure somehow this will benefit us all. I haven't worked with T2Well, but with ECO2N/M there are probably quite a few aspects that could be improved. I have at least missing wider options on porosity-permeability relation.
Looking forward to see your possible publications.
Godarzi
