Re: kinetic problem

[hui Tan <hut2+@xxxxxxxx>]

Dear Dr.Parkhurst,
         Here is another problem we met. The problem setup is:
100 meter 1d domain with 1 meter cells
5 m/ yr flow velocity
The aquifer initially have 0.8 mol/L K-fedlspar or 2% vol. The other 98% is 
quartz. The aquifer has 25% porosity
No initial values, but can precipitate gibbsite, kaolinite, and muscovite 
(K-mica)
The phase definition is taken from example 6 of phreeqc manual.

Gibbsite
               Al(OH)3 + 3 H+ = Al+3 + 3 H2O
               log_k          8.049

       Kaolinite
               Al2Si2O5(OH)4 + 6 H+ = H2O + 2 H4SiO4 + 2 Al+3
               log_k          5.708

       K-mica
               KAl3Si3O10(OH)2 + 10 H+ = 3 Al+3 + 3 H4SiO4 + K+
               log_k          12.970

       K-feldspar
               KAlSi3O8 + 4 H2O + 4 H+ = Al+3 + 3 H4SiO4 + K+
               log_k          0.875


The incoming solution has the following compositions

SOLUTION 		0
       units      mol/kgw
	  temp	 25
	  PH         8.1
       Al         1.e-13
       K          2.74e-5
       Si         9.25e-5
	  Cl         7.62e-5
	  Na         1.509e-4
	Ca         8.843e-4
	  Mg         1.09e-4

The K-feldspar dissolution rate is 10^(-12.5) mol/sec/sq meter.

The scenario is kaolinite has slow rate (K-feldspar rate /100),
muscovite = K-feldspar rate/10000


The problem we got is: If there is no transport part in the simulation, it 
worked well.After adding one dimensional  transport, the simulation can't 
converge.It warned that element Al and K has negative concentration in 
solution. Then it

stopped after bad RK step>500.


We attached the input files of transport and no-transport, so you can 
compare them.

Thank you very much.

Sincerely yours,
Hui Tan

Attachment: transport02.pqi
Description: Binary data

Attachment: Blackmesa-04-8.pqi
Description: Binary data