TMVOC-Temperature is out of range
Hello everyone.
I'm currently learning to use TMVOC, but I've encountered some problems. I've successfully simulated flow under only rainfall conditions, constant head, and isothermal conditions. However, I'm facing some unknown issues when simulating unsteady flow in a non-isothermal region. I'm using pumping wells to simulate unsteady flow.
Below is a description of the process.
I'm using TMVOC to simulate unsteady flow in a region. First, I created a flow model under constant head and isothermal conditions at this site. Then, under isothermal constant head conditions, I further added a certain number of pumping wells to obtain an unsteady flow model under isothermal conditions. Finally, I performed a non-isothermal simulation based on this, but again retained the pumping wells.
However, the situation shown in the image occurred during the simulation. Could you please provide some solutions? I've attached the relevant files.
I would like to express my sincere gratitude once again.
8 replies
-
The error messages indicate that the temperature has fallen outside the valid range for calculating the water saturation pressure (which is a function of temperature). If these messages occur only occasionally, they likely reflect a transient non-physical state during Newton iteration, where an intermediate solution produces an artificially low temperature. In such cases, the simulation can usually continue without affecting the final results. However, if the temperature at certain elements truly drops below 0 °C, this may indicate that excessive heat has been extracted from the system.
-
Dear Kenny, Thank you for your reply. This situation occurred during the simulation. I previously conducted a similar simulation, but with one-third the number of injection wells. This did not occur under those conditions. Furthermore, I did not manually interfere with the simulation. Could I interpret this as a threshold for the number of injection wells within a certain range? Or is there an error in my GENER setting? Could you please suggest specific improvement measures? Thank you . Best regards. Gao
-
Are you injecting with the correct fluid enthalpy?
Do T lower than O°C occur in injection elements?
-
Dear Alfredo Battistelli,
My current situation is as Kenny previously described: the simulation encountered negative temperatures midway through, but ultimately it still succeeded. I'm quite curious about this. Regarding the enthalpy issue you mentioned, I set the enthalpy to 25 degrees Celsius according to the site conditions, and the simulation results show no sub-zero grid lines. Could you please explain why this is happening? I look forward to your reply. Thank you sincerely.Attached is the relevant parameter setting for the pumping well using the DELV format.
-
As Kenny already wrote, during the NR iteration process the primary variables may go outside the working range for the EOS. Then a diagnostic message is printed and the time step is reduced.
In your case, as it seems there is no process forcing T to become less than 0, the run is able to go on.
I had some concern as the messages you show occured after a "reading multiphase flow input data" printout. But then you wrote the negative T messages were encountered during the simulation with injection wells.
I asked about enthalpy of injected fluid because that was a possible cause for negative T at injection elements. Enthalpy shall correspond to the desired injection T for the injected fluid, or fluid mixture. And it must be computed using the EOS correlations, not looking at some external source, as issues related to a different enthalpy reference state may be involved. But you finally obtained the expected results, so enthalpies were correct.
The limit about the number of sink & sources in the original stand alone TMVOC code is handled by the user through parameter MNOGN inside file T2.
Alfredo
-
Dear Alfredo Battistelli, thank you for your patient explanation. To be precise, the negative temperature occurred when I was setting up both pumping and injection wells simultaneously. However, this did not occur when I was only simulating atmospheric precipitation. What's puzzling is that this happens every time I simulate both pumping and injection simultaneously. Is this normal at present?
Best regards
Gao
-
A possible explanation is that when production and injection wells are in operation, reservoir conditions locally changes much more than during previous simulation with rainfall recharge only. Thus, it is more likely that some primary variable goes out of range, and this should not worry to much unless time step cuts become pathological.
This behavior may depend on the time step approach to. An 'aggressive' increase of time step allowing longer time steps may produce more convergence issues like primary variables out of range. As simulation time is proportional to the total number of iterations (including those for non converged time step) the optimum time stepping strategy should try to reduce the total number of NR iteration.
Alfredo
-
Dear Alfredo Battistelli, thank you for your patient explanation. I will adjust the parameters accordingly later.
Best regards
Gao
-
