Re: PHREEQC version 2.3

["David L Parkhurst" <dlpark@xxxxxxxx>]

The problem is with two As-F species, AsO3F-2 and HAsO3F-. The constants
for these species look much too strong to me. They imply that they account
for all of the arsenic, with the next species about 1e-10 smaller. The
convergence problem is caused because under most initial choices of log
activity F- and H2AsO4-2, these species cause all of the F to be calculated
to be in these species as well and the two mass balance equations are the
same on the left hand side within the numerical precision of the computer.
Right hand sides are related to the total concentrations, and these differ.

I would copy the definitions of these two species into your input file in a
SOLUTION_SPECIES data block. Then you can either eliminate them by
decreasing the log K's by 20, or you can decrease the log K's by perhaps 5
or 10, which will allow the program to converge. For the second case, the
distribution of species will be effectively the same as the original log
K's. The concentrations of the minor species will be larger, but with a
choice of K's, they will still be small relative to the As-F species. I
suspect there was a sign error or other tabulation error when they compiled
the data at LLNL.


David


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


                                                                                                                          
                      Michele Y Crouse                                                                                    
                                               To:      David L Parkhurst/WRD/USGS/DOI@xxxx                               
                      11/23/01 06:30           cc:      h2osoft@xxxxxxxx                                                  
In-Reply-To: <OF6A7E98F1.E00262C9-ON85256B0D.004A26AA@xxxxxxxx>
                      AM                       Subject: Re: PHREEQC version 2.3                                           
                                                                                                                          
                                                                                                                          




----- Forwarded by Michele Y Crouse/WRD/USGS/DOI on 11/23/01 08:29 a.m.
-----
                                                                                                                          
                      "RG&AJ Gerritse"                                                                                    
                      <geoprocc@xxxxxx         To:      <mcrouse@xxxxxxxx>                                                
                      net.au>                  cc:                                                                        
In-Reply-To: <OF6A7E98F1.E00262C9-ON85256B0D.004A26AA@xxxxxxxx>
                                               Subject: Re: PHREEQC version 2.3                                           
                      11/21/01 10:30                                                                                      
                      p.m.                                                                                                
                                                                                                                          
                                                                                                                          



With the attached input file for a recent groundwater sample I get a a
normal run with the WATEQ database, but with the LLNL database the program
will not converge. Taking As or F out makes the LLNL run converge. What can
the problem be?

regards
Robert Gerritse
Geoprocc Pty Ltd
Wembley Downs
Western Australia
(See attached file: Searl5.pqi)

Attachment: Searl5.pqi
Description: Binary data