Stopping in main due to flow nonconvergence

Hello everyone, 

I am using ECO2N to simulate CO2 injection into a formation with relatively low permeability. I am having trouble with an error popping up. Here is what I have done so far:

1. Pressure initialization (MOPR(1) = 2 in REACT) - ran this for 1000 years. Questions here: Does pressure initialization require a lot of time? Will my pressures in the mesh be different if I chose the same simulation time but different DELTMX for this case? 
2. Chemical initialization (MOPR(1) = 0 in REACT) - ran this for 10,000 years. I set up my DELTMX to be around 2 months (in seconds) since it saves computation time. 
3. Injection - I set up my GENER exactly similar to that of in P6-eco2n in the example problems. I decreased my DELTMX to 600 seconds. Changed SAVE and savechem to INCON and inchem. Made the times in INCON and inchem 0 to run the restart. When I run this simulation, after a couple of seconds I get this error:  Stopping in main due to flow nonconvergence.

I would appreciate any explanation as to what is happening. 

I have attached the three simulations I have done above for your reference. 

Kind regards,

Majid