
Updated:
26.10.2007
|
 |

GEM-Selektor version 2-PSI
DEFAULT CHEMICAL
THERMODYNAMIC DATABASE
Database Structure
Collaborators and Credits
The
GEM-Selektor v.2-PSI package is distributed with a default chemical
thermodynamic database that consists of three parts: (i) "kernel";
(ii) "complementary"; and (iii) "specific", selectable upon creation of
modelling
projects via GEMS "Elements" dialog.
(i) "Kernel"
part of GEMS default
database has been imported by T.Thoenen from Nagra/PSI chemical
thermodynamic data base 01/01 using the PMATCHC code. This well reviewed and documented database is
officially applicable to ambient conditions only (P=0.1 MPa, T=298.15
K). Selections of chemical elements, aqueous species, and
minerals in the
Nagra-PSI data base reflect the (rather specific) main purpose of its
compilation - support of geochemical modelling related to nuclear waste
disposal. In order to make the "kernel" database also usable at
elevated temperatures
and pressures, we have merged it with a subset of the SLOP98.DAT
file (http://epsc.wustl.edu/geopig),
as described in the unpublished technical report PSI TM-44-03-04. The
"kernel" GEMS version of Nagra-PSI data base can be used in most cases
for reliable calculation of aquatic equilibria up to 150 - 200 oC
at saturated vapor pressures.
(ii) "Complementary"
part of GEMS
default database consists of SLOP98.DAT
data directly imported into
DComp record format of GEMS.
The goal was to extend the potential number of GEM-Selektor users by
making it readily possible to calculate aqueous equilibria at
temperatures up to 1000 oC and pressures up to 5 kbar for a
much wider range of species and chemical elements than it is possible
with the Nagra/PSI database alone. Note that developers of the GEMS
package did not review this "complementary" data
and, hence, cannot assume any responsibility for its quality, even
though
the popular set of standard-state thermodynamic data SLOP98.DAT,
related
to the SUPCRT92 code and the revised
HKF EoS (Helgeson- Kirkham- Flowers
equation
of state), is at present a de facto standard, supplied in many
thermodynamic
modelling computer codes (e.g. ChemGeo, EQ3/6, GIBBS/Hch,
Selektor-C/W).
The HKF EoS subroutines for temperature/pressure corrections of partial
molal thermodynamic properties of aqueous species have been translated
from fortran
to C by M.Khodorivsky and incorporated into Selektor-A and then the
GEM-Selektor
code.
At
present, the "complementary" part of default GEMS database can be used
in
creating modelling project databases only together with the "kernel"
part. In this case, DComp and Phase records from both parts will be
copied into the project database files, but records from the "kernel"
part will be preferentially taken into calculation of chemical
equilibria. This is ho harm because the differences are almost
everywhere (except Eu species) quite small. However, the users who want
to calculate equilibria only using SLOP98 data, can do that after
deleting in the project database all records that are not marked "s9_"
or "t9_" in the last record key field. In
future
versions, it will be possible to create projects using the
"complementary"
data alone.
(iii) "Specific" part of
the
default database will contain thermodynamic data for specific
applications related e.g. to nuclear waste management, which extend the
"kernel" and/or "complementary" parts for minerals, solid solution
end-members, surface complexes and so
on. In the current version of GEMS-PSI, this part of default database
is
still under construction.
Please, note that
files of the default
GEMS database (located in the /program/DB.default directory)
are
write-protected, so that only developers can modify any data records
there. This is done
to ensure a backward compatibility between different modelling
projects.
A modelling application produced using GEMS is actually a directory of
database files, which, taken together, comprise a "modelling project".
All changes, input data, and results of modelling calculations
will be stored in the project database files, which can be modified by
the user who created it. A
project directory can be packed and shared with other users or contributed to this site as a test example.
The user can create any number of projects, but typically, such a
project is created for a separate geochemical modelling study in a
system of known elemental stoichiometry. Only part of the default
database (i.e.
species and phases defined by a selection of stoichiometry units) will
be
copied into the project database. The latter can be extended or
modified
any time at the user's discretion, but none of these changes will
affect the default GEMS database (unless the user will find and report
a bug in
a record from the default database).
The LES PSI
Thermodynamics Group
maintains the Nagra-PSI database provided as the default "kernel"
dataset with GEMS-PSI code and oriented mainly to specific nuclear
waste management applications. However, the GEM-Selektor code itself is
not application- or database-specific, so the main limitation of its
usage is the lack of internally consistent
input thermodynamic data. These can be compiled by other people,
converted
into GEM-Selektor format, and distributed as third-party chemical
thermodynamic databases (for instance, for hydrothermal, metamorphic,
igneous, metallurgical systems, and so on). Participation and
collaboration in this field is particularly welcome (see "Third-Party Contributions" file).
Notes about
the data imported from SLOP98.DAT
file
The original Slop98.dat file (downloaded from
http://epsc.wustl.edu/geopig) has first been split in seven ASCII files
in order to import separately
(i) aqueous inorganic species; (ii) aqueous organic species; (iii)
gases;
(iv) minerals without phase transitions; (v) minerals with one phase
transition; (vi) minerals with two phase transitions; and (vii)
remaining records. Each file was imported into DComp records of
appropriate configuration using a respective import script (available
in the GEMS import script library
as SDref records in the /DB.default/sdref.template.scripts.pdb
file):
"iscript-SUPCRT-aqs:0000:dcomp:"
"iscript-SUPCRT-gas:0000:dcomp:"
"iscript-SUPCRT-min1:0001:dcomp:"
"iscript-SUPCRT-min2:0002:dcomp:"
"iscript-SUPCRT-min3:0003:dcomp:"
During the data
import,
some formatting errors has been fixed in source files (the scripts read
the
input file as a stream and not in fixed field format as the
Supcrt92
program does).
At the second step, all DComp record keys (generated by import scripts)
were edited to bring them in accordance with the style and format of
the GEMS
default chemical thermodynamic database.
Finally, in every record, the chemical formula (DCform field) was
edited to bring it in conformance with the syntax of GEMS chemical
formulae. In
some cases, the explicit valence of redox-sensitive elements has been
entered to make sure that correct formula charge is calculated. Note
that the explicit
valences in chemical formulae have been entered solely to balance the
formula
charge and, hence, these numbers neither correspond to actual ionic
charges
nor reflect crystallochemical properties of ionic, covalent,
hydrogen
or other chemical bonds.
After that, each DComp record was re-calculated to check also the
internal
consistency of G0d, H0d, and S0d
values. In rare cases, where
the H0d value
has been found inconsistent, it was replaced by a consistent one
(computed
from G0d and S0d
values), and a remark about
this
correction was placed in the dsDval
field.
Data for minerals were
imported assuming the constant molar volume (independent of T,P).
For many
minerals with one or two phase transitions, the Slop98.dat file
contains
a "NULL" value (999999) in place of all or some parameters of the phase
transition; such values would result in errors when thermodynamic
properties are calculated in GEMS at temperatures above the phase
transition. Hence, those "NULL"
values are now provisionally replaced by zeros in the default database
file
of GEM-Selektor (will be available in the next released version);
please,
keep in mind that these data may still be incomplete and require
separate
checking before using them in calculations of chemical equilibria.
For a considerable number of minerals (e.g. zeolites), "NULL" values
(999999 cal/mol) are given for standard molar Gibbs energy and enthalpy
in the source SLOP98.DAT file, indicating that these values are
unknown. After import, these
data in GEMS default database have been replaced by 7777777 J/mol in
G0d[0]
and H0d[0] cells. All such minerals were not referenced in Phase
records to
prevent taking them into calculations of chemical equilibria.
Calculation of DComp record for NESQEHONITE resulted in Cpo298 =
-3689.18 J/K/mol, indicating an error in Majer-Kelly coefficients for
this mineral given in Slop98.dat file. For Pd-Oxyannite, the chemical
formula seems to be
wrong as it produces charge imbalance.
<to be extended>
References
Helgeson H.C., Kirkham D.H. and
Flowers G.C. (1981): Amer. J. Sci. v. 281, p. 1249-1516.
Johnson J.W., Oelkers E.H. and Helgeson H.C. (1992): Comput. Geosci.
v. 18, p. 899-947.
Shock E.L., Sassani D.C., Willis M. and Sverjensky D.A. (1997): Geoch.
Cosmoch. Acta v. 61, p. 907-950 (and references therein).
GEMS-PSI home
Last updated: 31. 03. 2003
© 2003 GEMS-PSI Development and Support Team.
|