Help with kinetic model for Mg2SiO4 dissolution

[ms14218]

Hello,

I am trying to model Mg2SiO4 dissolution using the llnl.dat at different temperature and pressure conditions. However, the rates script which i am using (adapted from Zhang et al., 2019 (A library of BASIC scripts of reaction rates for geochemical modeling using phreeqc), seems to provide reaction rates that are far too rapid. I would like to understand if there is a missing piece of information or error in my script, which would explain this. Or what alternatives there are to provide a model which is more consistent with the slower dissolution of the sample which was confirmed in experiments (the sample was nowhere near entirely leached of Mg within an hour as the model suggests).

Second issue: once I change the timespan of the dissolution, the data (displayed using the USER_GRAPH function) changes, i.e. I run a 1 hour model and it will show a faster increase in Mg release with temperature. However, if I run the same script but for a 10 hour dissolution period, the model will show different results, with an optimal Mg release rate achieved somewhere in the middle of my temperature range (185C) rather than at the maximum selected temperature (250C). I have tried to tweak the script to fix this issue but to no use.

Help regarding these issues would be greatly appreciated.

[dlparkhurst]

You do not give enough information to know the solution, temperature (and pressure) where the rate is too fast. Using a carbonate groundwater, it takes a couple hundred years to reach equilibrium at 25 C, which seems plausible. Palandri and Kharakha give a temperature range of 25 to 65 C, so maybe you are outside the range of applicability of their rate equation and will need to develop your own parameters.

Note that llnl.dat accounts for pressure implicitly in the database log Ks. It uses 1 atm until 100 C and then the steam curve above 100 C. phreeqc.dat has explicit, independent definitions of temperature and pressure.

You have two parameters in the KINETICS definition that affect the rate directly. Decreasing either one of those will slow the rates.

You will have to be more specific on your second issue for me to understand what you are talking about. You will get a different average rate depending the length of your simulation. Using INCREMENTAL_REACTIONS and the following USER_GRAPH will allow you to plot the rate over each integration interval. Note that the cumulative time for integration is not the same as when INCREMENTAL_REACTIONS is not used.

Code: [Select]

USER_GRAPH 1
    -headings               years Rate
    -axis_titles            "Years" "Rate, mol/s" ""
    -axis_scale x_axis      auto auto auto auto log
    -initial_solutions      false
    -connect_simulations    true
    -plot_concentration_vs  x
  -start
10 GRAPH_X TOTAL_TIME
20 GRAPH_Y KIN_DELTA("Forsterite")/KIN_TIME
  -end