9.4 Problem No.4- Five-spot Geothermal Production/Injection(EOS2)
Dear all,
I have been studying and trying to run this problem example in the manual cause it is similar to my Project. I'm trying to view the simulation results using TOughViewer. When i upload the MINC file, then comes '' BlockName not found in CONNE table'' and when i click on ''OK'' then comes info if not wrong, about the first continuum connection, in this case '' fracture''. this happens also With another MINC file for a different simulation, can someone tell me what all this about ?
Secondly when i upload the output file, then comes "BALLA problem.Double..." error, and when i click on "OK" then comes the of different Components and phases,(GAS, WATER, CO2 etc..) and this error Message between them. The final result is also Attached, With no information about Nodes coordinates. Then, i end up not viewing the result. All the files are Attached ( Error files, input file, and the output file) . How can can fix this problems and view the result ??
Regard
Nathan
22 replies
-
Nathan,
I recommend you contact the developers of TOUGH Viewer directly, as I'm not sure they monitor this Forum.
In general, none of the visualization packages I know can directly process MINC files, simply because MINC has overlapping continua, i.e., multiple grid blocks with the same coordinates. It is not even clear to me what you would want to visualize if you loaded in the entire file...
I usually do a little pre-processing step, in which I split the MINC file into separate MESH files, one for each continuum - then I visualize pressures, flow rates, etc. for each continuum separately.
Stefan
-
Dear Stafan,
I am still struggling With the same problem,
The error massage has disapeared after i changed the output ''MOP'', but it seams i don't have a Clear Picture of what i am doing.
My wish has been to design the injection of CO2 into a geothermal system and visualize the reservoir pressure and heat Development and Production of heat at Production well With time. basically, the model should be a fractured reservoir, but since you advised me to work With each continuum at a time, i decided to run the problem without ''MINC'' first, thus working With primary mesh only, then i will work With the ''MINC'' after
I have tried to modified the problem in the TOUGH manual, but i only get constant pressure and temperature for all elements.
Please what should i do to get varied pressure and temperature With time ?
sencondly, what should i do to invoke Production well data (temp or heat Production With time) ?
enclosed is my input and out put file ..
Regard
Nathan
-
Nathan,
I apologize if I have not been clear with my suggestion: I meant to say that you should separate the continua when visualizing the MINC results, not when setting up and running the problem (you should run the full MINC model)!
Your input file is a mix between EOS1 and EOS2. Also check MOP(16) (should be 4, not 0, to make time steps increase; it seems MOPs are shifted to the right by one slot), MESHM (delete), GENER (there is no COM3 in EOS2; for production, use MASS), PARAM.4 (check primary variables for EOS2), etc. - they all have issues. Please follow the instructions in the manual for specifying time-dependent injection or production rates.
If somebody has time to go through all these blocks with Nathan in more detail, please chime in!
Stefan
-
Stefan Finsterle
Dear stefan !
I have modified the input file according to Your instruction.
enclosed is the input fil too, you can have a look. pls alert me in case of anything not appropriate. The problem is not seperating the continua in order to visualize the temp and pressure. an example of seperated 1st continuum which is the fracture is pasted in the MESH file, and when i oppload it in a prosessor, the prosessor can not read it, just Wonder if i have pasted it wrongly or may be the format is not appropriate ?
Regard
Nathan
-
Nathan,
Please look at your own MESH file - it is obviously not properly formatted (please check the manual for proper formatting of blocks ELEME and CONNE, and be aware of empty lines).
You could actually just take your primary mesh, put it in a MESH file, and run EXT.
Stefan
-
Dear Stefan,
I have been running this simulation With the Maximum time of 30 years, from the output i only get the result of almost 2 years which is the Maximum convergence time of the simulation. How do i extend the convergence time and get the result of maybe 25-30 years ?
Nathan
-
Nathan,
Please look at the results in detail at the time the simulation stops. Recall what the system looks like, and what you want to do. It is clear to me that the way it was set up, it will eventually crash (i.e., it does not make much sense to try to force it through the desired end time).
Whenever there is a convergence like that, in many (but not all) cases this is just an indication that TOUGH2 is asked to simulate something that is unphysical.
I hope you understand that I cannot rerun all models that are posted on the Forum or sent to me directly - I simply do not have the time and resources to do so. Please try to understand the system, explain what changes you made to remedy the problem, what worked and what did not, and then ask a specific question, and I will be happy to try to answer the question to the best of my abilities.
Best,
Stefan
-
Stefan !!
i think i understand the principle behind the software now,. My wish has been to try to modell injection of CO2 into a geothermal reservoir and study the heat, T and P changes using (EOS2). so far i have tried that, but i keep getting constant temperature, the reason behind i have failed to figure out for almost a week With intense working,
secondly, how can i specify a reservoir intially saturated With CO2, i tried that in the PARAM Block by varying the percent saturation of gas, but i keep getting more water than CO2 in the out put file when i look at mass in Place when Placed initial time step = 0. ??
The Third question , when injecting CO2 With lower temp than in reservoir, the value of the enthalpy in (GENER Block ) should be the same as in Reservoir or should it be calculated from inj. temp and reservoir pressure ??
Pls Stefan, i am trying to reduce asking you frequent question by working hard, but this time i have failed to figure these Three questions after working for many many hours With them. enclosed is my input.
Best regard.
-
Nathan B.
Dear Nathan,
sorry if I step into this conversation. In your case start from a very simple case, too many things could be wrong...
Can you run the 5-well-problem of the TOUGH2 manual (problem n. 4) and get identical results to those of the manual?
Once you get consistent results with problem n. 4, make the changes you need to match your specific problem.
Andrea
-
Nathan,
Please follow Andrea's advice. Also check the manual, specifically the sample problems. You have to strictly adhere to the TOUGH formatting conventions (e.g., real numbers need a decimal point; see your GENER block!).
(1) Your temperature is (almost) constant most likely because your injection rate is probably off by a factor of 1E4 (because of formatting error mentioned above).
(2) Changing X(2) in PARAM.4 works just fine to change initial gas saturation.
(3) You cannot specify a time step of zero. Set MOP(7)=9 to get initial conditions.
(4) Specify the value of the enthalpy that corresponds to the desired temperature of the CO2, or inject at a predefined temperature following Karsten's instructions , which you can find on the Forum at http://esd1.lbl.gov/files/research/projects/tough/support/02-pruess.pdf
Hope this helps,
Stefan
-
Stefan Finsterle
Stefan,
My expectation of the result has been that the reservoir temperature will increase from the injection to Production at a specific time / in overall drop With time. Instead i am still getting almost constant temperature, and the temperature is building up at injection point. What could be the reason behind this ? fail to understand. Hope this could be my last question.
Nathan
-
Nathan B. did you check your sign in GENER?
-
Andrea Borgia
Yes i did Andrea Borgia . Temperature is changing but not that big variation, and it is building up on injection block. Enclosed is my input, try to run it as well
Thanks
-
Nathan B.
Nathan,
can you compare your Input file to that of problem 4? What do you find different? Can you run correctly the exact problem 4?
-
Andrea Borgia
Hi Andrea,
Your were right, i run the problem With injection of COM2, just as suggested here by Stefan; https://tough.forumbee.com/t/h48y30/simulating-co2-injection-using-eos2. it gave me the same trend of results as example No.4. but the results are not in line With the literatures, temperature should be increasing from the injection to the Production, but the example problem No.4 and my simulation gives the opposite results, thus ( temperature decrease from injection to Production ), the good trend comes when i use either MASS or WATE as an injection fluid even With CO2 Thermo reservoir properties, which i should not use, because I am working With CO2 as inj. fluid. I haven't manage to figure the problem, since you last wrote to me. Try to look at the input i am using, maybe you will manage.
Regard
Nathan
-
Nathan B.
Nathan,
What EOS module are you using?
Andrea
-
Andrea Borgia
I am using EOS2,
Nathan
-
Can you send me the output file from this input, please?
-
Andrea Borgia
Here it is
-
Nathan,
I see a number of things that can perhaps create problems.
In the PARAM block you have defined the primary variables for EOS2 as:
150. 0.001 4.90E5
If you read in the tough2 manual how to define primary variables for EOS2 (page 34), you may observe that the values you have used for the primary variable appear to be reasonable for EOS1 (page 30) which is the EOS module used in example 4, but the same numbers cannot be use for EOS2 that requires different primary variables. In particular the partial pressure of CO2 instead of the mass fraction of water2.
If I where you, I would start again. Trying to achieve first a simple equilibrium solution. 1) Run the problem without the GENER block with only water (one phase) at a given temperature and a very small partial pressure of CO2. If you are satisfied with the equilibrium conditions that you have achieved (that is pressure/temperature/CO2 partial pressure), then
2) Re-run the problem increasing the partial pressure of CO2. You may eventually reach equilibrium conditions that are 2-phase.
3) Always look at the SAVE file and observe how the primary variables are re-assigned to the various cells for 1- or 2-phase conditions. Compare these variables to the definitions given for EOS2.
4) Finally, once you are satisfied with the equilibrium results (don't take the reaching of equilibrium conditions as a dummy result, you will have to take this step probably for every other tough2 simulation you will ever do) than add the GENER block and try to run it again.
I would also use a very small permeability for the matrix (1.E-20) in the block ROCKS as opposed to zero. In general TOUHG2 does not like zeros.
Also, for linearity, I would not put blank lines in the input file unless they are indicated as necessary by the manual in any specific block.
A.
-
Dear Andrea / Stefan !
As you both t knows I'm working with CO2 injection using EOS2. I am Puzzled again,
I do manage to achieve the equilibrium for single phase liquid (water & dissolved CO2). My problem is the transition to two phase. From the manual page 32, it says that initialization of two phase from a liquid is done by putting Sg = 0.000001 and Pg = Psat. I have tried this but did not succeed to reach the equilibrium in two phase, my question is ; Psat required is it for water or CO2 when injecting CO2 in a reservoir initially saturated with water? My initial condition is P = 290 Mpa and T= 100 C , for which CO2 is in the Supercrtical state at this point? In short how to Initialize a two-phase condition for CO2 & water reservoir for which CO2 will be at Supercritical state when injected ??
Regard
Nathan -
Nathan B.
Nathan,
Page 32 is for EOS1, not EOS2! Check Table 8 and you will see that two-phase conditions are initialized by specifying Pg (or T), Sg, and PCO2 as the three primary variables. Sg is the saturation of the gas phase, which consists of CO2 and water vapor. So the difference between Pg and PCO2 is Psat(T) and thus determines T (or, if you chose T as the first primary variable, determines Pg). Initializing EOS2 is indeed tricky.
However, you said that you wanted to inject CO2 into an initially liquid saturated reservoir. Do just that, i.e., initialize your system as single-phase liquid and specify a CO2 mass injection rate in GENER, and TOUGH2 will do the phase changes by itself if the thermodynamics call for a phase change.
Or do it as in sample problem rfp, which actually initializes the reservoir with 1% gas in it (the gas consists mostly of steam, with about 10% of CO2).
Stefan