How to fix this error: N-R didn't converge after 999 its., maximum residual = 1, Xi = 0.0000

[kevinning51]

When I am trying to run simulations of water evaporation, reaction of groundwater and mineral phase (among others) I keep getting this error message:

 

N-R didn't converge after 999 its., maximum residual = 1, Xi = 0.0000

 

I tried to make the simulation as simple as possible, but the software still indicate errors.

 

the script is as following:

> data = "C:\Program Files\Gwb\Gtdata\thermo.dat" verify

> time start = 0 day, end = 1 yr

> length = 20 cm

> left = inlet

> right = outlet

> Nx = 20

> discharge = 5 m/yr

> porosity = .3

> scope = initial

> H2O = 1 free kg

> Na+ = 0 mg/kg

> Mg++ = 0 mg/kg

> Ca++ = 0 mg/kg

> K+ = 0 mg/kg

> balance on Cl-

> Cl- = 0 mg/kg

> SO4-- = 0 mg/kg

> HCO3- = 0 mg/kg

> pH = 7

> scope = inlet

> H2O = 1 free kg

> Na+ = 1585 mg/kg

> Mg++ = 176.5 mg/kg

> Ca++ = 50 mg/kg

> K+ = 46 mg/kg

> balance on Cl-

> Cl- = 2300 mg/kg

> SO4-- = 320 mg/kg

> HCO3- = 14.2 mg/kg

> pH = 8.2

> TDS = 45000

> react -3e3 kg/m3/yr of H2O

[Brian Farrell]

Hello,

 

I think your problem arises from the 0 values for concentration that you specify in the Initial system. If these indeed are supposed to be 0 in the real world, they should be replaced with negligibly small values in your thermodynamic model. You should take a look at some of the examples in the Reactive Transport Modeling Guide. For example, in 3.9: Pb contamination, clean water flushes a Pb contaminated aquifer, but even the clean water has an exceedingly small but nonzero value for Pb concentration.

 

Hope this helps,

 

Brian Farrell

Aqueous Solutions LLC

[kevinning51]

Yes. It is because of the zeros.

Thanks for the reply.