webmaster Posted December 26, 2004 Share Posted December 26, 2004 From: Paul Barton Subject: Database If I run a very simple system such as 10 mg/kg Al+++, 300 mg/kg SiO2(aq), 300 mg/kg Na+, and 500 mg/kg Cl- and a pH of say 6, using REACT, and with the current best data base (thermo.com.V8.R6.full), the REACT_OUTPUT file tries to tell me that quartz plus (depending on the temperature) either Diaspore or Corundum is stable. That's an absurd outcome. What about Kaolinite, or Pyrophyllite, or Andalusite, etc.? Old data bases do not have this problem, but that is faint praise indeed. I am wondering whether (wherever thermo.com.V8.R6.full gestates in the entrails of LLNL) only a partial reassessment of the consequences of revising Al chemistry by Pokrovskii and Helgeson (1995) has been made. A change in log Ko for Al+++ should impact all Al compounds and species since Al+++ is the basis species. But simply making an adjustment, ( based, for example, on the shifts in log Ko for Corundum ) runs the risk of making extra corrections where they have already been applied. From: Craig Bethke Subject: Re: Database I think you are confusing bigger with better. The V8-R6 database is the largest available, and perhaps this led you to think of it as the “current best data base�. In fact, large thermo databases by nature contain data that range widely in quality, and tend to suffer in terms of internal consistency. This database has important attributes -- a broad range of organic species, for example, and lots a data for radionucleides -- but also some widely appreciated problems. In terms of consistency between aluminous and aluminosilicate species, I think you will find another database such as thermo.dat more suitable. Remember that the thermo dataset is input to a model, and choosing appropriate input data is the user's responsibility. Fortunately, you need never treat a thermo database as a black box, since you can quickly use Act2 to plot the relations among aqueous species, minerals, and gases. In this way, you can verify before you begin modeling that the dataset is adequate in scope and sufficiently consistent to describe your system at the conditions of interest. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.