Close to surface simulations, low ionic strength

[zsszabo86]

Hi!
I think I would use llnl.dat, but I need a verification for what is the best database to apply for river waters reacting with surrounding rock.
Many thanks in advance
Zsuzsanna

[dlparkhurst]

It depends mostly on what elements and minerals you need. llnl.dat is the most extensive, but it also has some unusual valence states for elements and some polyatomic species that can cause convergence problems. Amm.dat, phreeqc.dat and pitzer.dat have additional to calculate specific conductance and density, and have been fit to higher temperature and pressure. phreeqc.dat is a subset of wateq4f.dat, although wateq4f.dat does not have the specific conductance and density parameters. pitzer.dat is good for high ionic strength, but has a limited set of elements. Amm.dat is the same as phreeqc.dat, except ammonia has been split into a separate element to avoid automatic equilibrium with other nitrogen redox states.

So take your pick. If you are modeling alumino-silicate reactions, you are going to have big uncertainties in any event. The primary species for Si is SiO2 in llnl.dat and H4SiO4 ina all other databases, so a conversion is necessary to use a mineral from llnl.dat in another database. I think you are probably choosing among llnl.dat, wateq4f.dat and phreeqc.dat (or Amm.dat).