Subroutine NSIN

Hi everyone,

I am running ECO2N in Tough2-mp, to simulate CO2 injection in a 3D model. Initialization of the model goes without problems, and the output is used as initial conditions for the CO2 injection. The model simulates injection for 10-20 days (it varies between runs) but then it stops without reaching the desired simulation time (1 year). Output shows the following error:

 ### no convergence in NSIN at * 1379   * for (T, XS, XLCO2) = ( 75.7524, 0.150000E+00, 0.100499E-05) P00=0.695328-309
     latest parameters are   PS=         NaN   X2S=         NaN
 +++++++++   CANNOT FIND PARAMETERS AT ELEMENT * 1379   *          XX(M) = 0.201133E+08 0.150000E+00 0.994484E-06 0.757524E+02
 +++++++++ REDUCE TIME STEP AT (   35, 1) ++++++++ ++++++++   NEW DELT =0.192217E+04

This same message shows for some other  cells. I've checked the output for those cells and no NaN or strange values shows until the error occurs.

The mesh is refined around the injection well, but if the model is run without mesh refinement (for instance all cells with same sizes 200 m x 200 m x 10 m), the simulation runs until the desired 1 year .

Does anyone has any idea with this occurs and how to solve it?

Thank you,

Júlio