File: RELEASE.TXT MINTEQA2 Model System Release Notes version 4.03 May 2006 Center for Exposure Assessment Modeling (CEAM) National Exposure Research Laboratory - Ecosystems Research Division Office of Research and Development (ORD) U.S. Environmental Protection Agency (U.S. EPA) 960 College Station Road Athens, Georgia 30605-2700 706/355-8400 _____________________________________________________________________________ CORRECTIONS AND IMPROVEMENTS TO MINTEQA2 VERSION 4.02 ____________________________________________________________________________ New versions of MINTEQA2, PRODEFA2, FRMT and UNFRMT were generated for modern operating systems (Win2K/XP) due to the failure of the Lahey compiled versions to correctly operate on these systems. Intel FORTRAN v9.0 was used to recompile the codes with attendant minor changes in operation in MINTEQA2 and PRODEFA2. The new versions will run on all MS Windows versions. The programs are now run directly; the DOS batch files that launched the programs have been removed. To initiate a MINTEQA2 simulation enter the runtime image's name; "MINTEQA2". PRODEFA2, FRMT and UNFRMT all operate in a similar fashion. ----------------------------------------------------------------------------- ANSI extended codes that perform screen manipulations were removed from MINTEQA2 and PRODEFA2 due to the complexity of loading the various versions of the ANSI.SYS device driver for the differing command interpreters (COMMAND.COM, CMD.EXE) in the various operating systems. Screen IO for these programs is now sequentially generated and scrolls in the normal manner for command interpreters. ----------------------------------------------------------------------------- The new version of MINTEQA2 requires versions of TYPE6.UNF and THERMO.UNF generated by the new Intel compiled version of UNFRMT. If users have their own customized versions of these databases, they should first run the new FRMT program and then the new UNFRMT version to generate Intel FORTRAN direct access versions of these files. ----------------------------------------------------------------------------- A new component database (MMHG&TBT.DBS ) containing Monomethylmercury (MMHg) and Tributyltin (TBT) is distributed with this version of MINTEQA2. A testing input file (MMHG&TBT.INP) and comparison output file (test_output\MMHG&TBT.OUT) are included. To run the test users can: 1) Rename the COMP.DBS to save it and rename the MMHG&TBT.DBS to COMP.DBS; then run the model. Or, 2) Edit the DBASE.INI file directly and change the line “COMPONENT DATABASE: comp.dbs” to “COMPONENT DATABASE: mmhg&tbt.dbs”; then run the model. Or, 3) Run PRODEFA2 to generate the correct DBASE.INI file prior to running the model. Refer to the "MMHG&TBT.pdf" paper in the \docs directory for further information. ----------------------------------------------------------------------------- No direct changes to the science simulated by the model have been made but users should note that, to date, small changes (on the order of 10e-19) occur in some numerical mass balance results generated by MINTEQA2 executing the testing input files. No programmatic reason for this discrepancy has yet been found and no scientific assessment of the significance of these differences has been obtained. The differences may be due to accumulated rounding errors for numerical operations or compilation switch incompatibility but this has not been ascertained at this time. Testing consists of running the newly compiled MINTEQA2 code with the input files provided and comparing the results generated with the standard test results contained in the test_output subdirectory. Current testing results are: test1a - passes test1b - passes test2a - fails test2b - fails test2c - fails test3 - passes test4 - passes test5a - passes test5b - fails test5c - passes test6 - passes test30 - fails The tests fail in the following way; all differences occur in Part 3 of the output files in "Parameters of the component most out of balance" section. test pgm variable component_id/name results standard results ------ ------------ ----------------- ---------- ---------------- test2a array t 330/H+1 -7.200e-20 -7.115e-20 test2b array t 330/H+1 -1.050e-19 -1.042e-19 test2c array t 330/H+1 -7.200e-20 -7.115e-20 test5b array y 330/H+1 0 -4.337e-19 test30 array t 330/H+1 0 -1.158e-21 test30 array t 2/H2O 0 -1.158e-21 test30 array y 580/PO4-3 0 1.548e-21 "array t" holds the total concentration values, "array y" holds the residual values, "results" are what is generated with the newly compiled code, "standard results" are numbers from the comparison files in the test_output directory. The numbers in question on the reports are generated in the SOLIDX routine inside nested loops; the total concentration and residual storage arrays are declared and sized in the MINTEQA2 include file. -----------------------------------------------------------------------------