X1t Flow through sandstone

[Anastasia]

Hello,

 

I'm constructing a 1D reactive transport model where fluid flows through sandstone. I consistently get this (or, similar) message:

 

Solving for composition of inlet fluid.

Solving for initial state of each nodal block.
Residuals too large, 652-th interation
Largest residual(s):
Resid Resid/Totmol Cbasis
--------------------------------------------------------
Cl- 190.2 1.207e+200 1.031e-196
--------------------------------------------------------

 

I ran SpecE8 on both initial and inlet fluid compositions, and they both are about charge balanced:

* Initial: Charge imbalance = -0.012 meq/kg (-0.001151% error);

* inlet: Charge imbalance = -0.005351 meq/kg (-0.0005109% error)

 

I also ran React on the inlet fluid composition reacting with quartz, calcite, K-feldspar, and hematite, and this calculation has converged.

 

Please let me know if you have any ideas of how to set up this 1D reactive flow model.

 

Thank you,

Anastasia

[Brian Farrell]

Hi Anastasia,

 

Can you attach your X1t script please?

 

Thanks,

Brian Farrell

Aqueous Solutions

[Anastasia]

Of course! Its attached.

 

Thank you so much for all your help,

Anastasia

[Anastasia]

o.k. now it's attached

Sandstone.x1t

[Brian Farrell]

Hi Anastasia,

 

Are you sure the constraints you've specified (concentration values and units) for the Initial system and Inlet fluid are correct? The K+, for example, is .00362 mg/l in the Initial system and .00362 molal in the Inlet fluid. Perhaps this is just a coincidence, but maybe one of the values or units is wrong. You might also think about your conceptual model of the system. For example, you've specified that both the Initial and Inlet fluids are in equilibrium with Quartz and K-feldspar, but you've also set kinetic rate laws for these minerals. Are you sure that both of these fluids are supposed to be in equilibrium with each other?

 

On a different note, it's sometimes easier to converge upon a solution when you charge balance on a cation than on an anion. In your fluid, if balancing on Cl- fails, you might try Na+.

 

Regards,

Brian

[Anastasia]

Brian,

 

Thank you for getting back to me. Good point about equilibrium with the minerals and kinetic rate laws. I'll change that. I'' double check the concentrations and units too. The conceptual model is to react a sandstone (with initial brine) with the similar brine when it has more CO2 dissolved in it.

 

Do you have a suggestion of what variables can be changed if a reactive flow model runs for a fraction of time (a few years), then crashes with "N-R didn't converge" message shown below?

 

Thank you so much for all your help, I really appreciate it!

Anastasia

 

Message:

-- Warning: Reactants at node 0 are not charge balanced, z = 0.506328
*N-R didn't converge after 999 its., maximum residual = 19.1, Xi = 0.3137

N-R failed at node 0
Attempting to write last good time step for plotting.

-- Can't converge, abandoning simulation.
-- Warning, Node 0 has a charge imbalance of -0.2454 faradays