CONVERGENCE FAILURE and STOP EXECUTION AFTER NEXT TIME STEP

[Prince Kumar]

Hello, I am facing the following error.

I am new on PetraSim. If there is any solution please let me know. Thanking You 

 

 1C 2(****,  2) ST = 0.841700E+09 DT = 0.800000E+03 DX1= 0.730626E+00 DX2= -.341338E-11 T =  34.953 P = 10572413. S = 0.254903E-03
 +++++++++ REDUCE TIME STEP AT (****,  9) ++++++++++++++++   NEW DELT = 0.400000E+03
  1C14(****,  7) ST = 0.841700E+09 DT = 0.400000E+03 DX1= 0.397600E+00 DX2= 0.489379E-12 T =  34.194 P = 10329049. S = 0.571449E-04
 +++++++++ REDUCE TIME STEP AT (****,  9) ++++++++++++++++   NEW DELT = 0.100000E+03
  1510(****,  1) ST = 0.841700E+09 DT = 0.100000E+03 DX1= 0.000000E+00 DX2= 0.000000E+00 T =  33.989 P = 10095439. S = 0.000000E+00
  1B64(****,  4) ST = 0.841701E+09 DT = 0.200000E+03 DX1= 0.183087E+00 DX2= 0.452424E-12 T =  35.054 P = 10586076. S = 0.268510E-03
  1B61(****,  4) ST = 0.841701E+09 DT = 0.200000E+03 DX1= 0.183087E+00 DX2= 0.452895E-12 T =  35.054 P = 10586077. S = 0.268512E-03
 +++++++++ REDUCE TIME STEP AT (****,  9) ++++++++++++++++   NEW DELT = 0.500000E+02
  1510(****,  1) ST = 0.841701E+09 DT = 0.500000E+02 DX1= 0.000000E+00 DX2= 0.000000E+00 T =  33.989 P = 10095440. S = 0.000000E+00
  1510(****,  1) ST = 0.841701E+09 DT = 0.100000E+03 DX1= 0.000000E+00 DX2= 0.000000E+00 T =  33.989 P = 10095440. S = 0.000000E+00

 +++++++++ CONVERGENCE FAILURE ON TIME STEP #**** WITH DT = 0.200000E+03 SECONDS, FOLLOWING TWO STEPS THAT CONVERGED ON ITER = 1
           STOP EXECUTION AFTER NEXT TIME STEP
 +++++++++ REDUCE TIME STEP AT (****,  9) ++++++++++++++++   NEW DELT = 0.500000E+02
  1510(****,  1) ST = 0.841701E+09 DT = 0.500000E+02 DX1= 0.000000E+00 DX2= 0.000000E+00 T =  33.989 P = 10095440. S = 0.000000E+00

 

[Alison]

Prince,

 

Unfortunately, its very hard to determine what is causing this by looking at this log alone.  Convergence problems like this are not uncommon in simulations where there are phase changes.  Here are some very general suggestions:

 

1.  You could try restarting your model by loading the SAVE file as initial conditions, and changing the RE1 (Relative Error Criterion) to a smaller value.

2.  You could experiment with modifying your RP and CP curves.

3.  If you are modeling as injection, you could try initializing your model with a very small amount of gas already present.

 

If you continue to have problems, you are welcome to send your SIM file to tech@rockware.com and we'll see if we can provide additional suggestions.

 

Thanks,

Alison

[Prince Kumar]

Thank You. I will try it. 

[Prince Kumar]

Hi Alison, 

Again, I am getting a convergence problem in another simulation. The following error is shown as: 

 

 END OF TOUGH2 INPUT JOB --- ELAPSED TIME =   1.3120 SECONDS
  1B11(   1,  1) ST = 0.550511E+10 DT = 0.100000E+03 DX1= 0.000000E+00 DX2= 0.000000E+00 T =  33.949 P = 23141026. S = 0.000000E+00
 +++++++++ REDUCE TIME STEP AT (   2,  9) ++++++++++++++++   NEW DELT = 0.500000E+02
  1B11(   2,  1) ST = 0.550511E+10 DT = 0.500000E+02 DX1= 0.000000E+00 DX2= 0.000000E+00 T =  33.949 P = 23141026. S = 0.000000E+00
  1B11(   3,  1) ST = 0.550511E+10 DT = 0.100000E+03 DX1= 0.000000E+00 DX2= 0.000000E+00 T =  33.949 P = 23141026. S = 0.000000E+00
 +++++++++ CONVERGENCE FAILURE ON TIME STEP #   4 WITH DT = 0.200000E+03 SECONDS, FOLLOWING TWO STEPS THAT CONVERGED ON ITER = 1
                            STOP EXECUTION AFTER NEXT TIME STEP
 +++++++++ REDUCE TIME STEP AT (   4,  9) ++++++++++++++++   NEW DELT = 0.500000E+02
  1B11(   4,  1) ST = 0.550511E+10 DT = 0.500000E+02 DX1= 0.000000E+00 DX2= 0.000000E+00 T =  33.949 P = 23141026. S = 0.000000E+00
 END OF TOUGH2 SIMULATION RUN --- ELAPSED TIME =     4.125 SEC-- CALCULATION TIME =     2.813 SEC-- DATA INPUT TIME =    1.312 SEC
 

Could you please help?

Thank You 

[Alison]

Prince,

 

There seems to be a problem with element 1B11.  I would suggest that you take a look at the location of that element relative to injection cells and also to the boundary between various phase states.  Particularly with ECO2N (ECO2M also, I suspect), you can run into convergence problems at the boundary between, for example, injected CO2 gas and the aqueous phase.  The solution to this is often related to changes to RP or CP curves.

 

If you continue to have problems, please send your SIM file to tech@rockware.com and we'll take a look.

 

Best Regards,

Alison