RE: convergence problem

[Henning.Prommer@xxxxxxxx]

Hi,

thanks a lot for your quick reply.
As the problem occurs during a transport 
simulation it seems the best if I used
the last option you proposed.
When writing the input for the 
reaction step I will check if the O(0) conc
is 0 and then set it to a small value ... or
do you think it would be a problem if I generally 
avoided O(0) to be 0 ?
As far as I remember I previously fixed a problem
by replacing very small concentrations (eg 1e-35)
with zeros in the input file(s).

Cheers
   Henning



-----Original Message-----
From: David L Parkhurst
To: Henning.Prommer@xxxxxxxx
Sent: 8/30/2002 11:08 PM
In-Reply-To: <OF27BACA59.8AD84324-ON87256C25.00523F9E@xxxxxxxx>
Subject: Re: convergence problem


> I have come across a PHREEQC convergence problem for which I couldn't
find a cure. I am not sure, maybe there is something wrong with
my input or my definitions in the database. Could you, if you have
the time, have a quick look at it ? I have attached the input file
and the database file that I used.
I had this type of problem before after I included Ferrihydrite or
Goethite
as equ. mineral (allowed precipitation) in an anaerobic, pyritic aquifer
in
which aerorobic water was injected.

Pretty subtle problem, but I think it is caused by fixing the pe but
letting the pH vary. You are right at the edge of water stability and
the
pH must increase in this problem to produce charge balance. The result
is,
in a sense, stepping over the stability of water boundary. The real
effect
is that the dissolved oxygen concentration is calculated to be large.
This
in turn affects the activity of water, and the numerical solution gets
unstable. Maybe the numerical solution could be better, but the real
problem is that you are calculating a higher pH than input with a fixed
pe.

Several things will alleviate the problem: (1) if you do not charge
balance
you should have no problem, (2) if you charge balance on anything but pH
and pe, (3) you set the pe by a redox couple, or (4) define the
concentration of O(0). The last may be the easiest, if you set O(0) to a
small number, say 1e-18 things will probably work ok.

David


David Parkhurst (dlpark@xxxxxxxx)
U.S. Geological Survey
Box 25046, MS 413
Denver Federal Center
Denver, CO 80225

Project web page: https://wwwbrr.cr.usgs.gov/projects/GWC_coupled