STIMAX parameter: Which ionic strength is considered?
I have a question concerning the STIMAX parameter in the solute.inp file in record 3. As described in the TOUGHREACT V3.0 manual, STIMAX acts as:
"Geochemical calculations are skipped at grid blocks where the stoichiometric ionic
strength is more than STIMAX."
This is of course useful for salt dominated brines, where the primary aqueous components are usually ions. Hence, the stoichiometric ionic strength should be generally larger than the normal ionic strength. However, this does not hold for high concentration dissolution of acid impurities included in CO2 gas phase, where the corresponding aqueous primary components are neutral (SO2(aq), HNO3(aq)) and the prevailing secondary species are ionic.
I have done some simulations with TOUGHREACT V3.0 / ECO2N by varying the STIMAX parameter between 3, 6, and 8. I manually computed both ionic strengths from the time.out file from one particular cell in a 1D radial symmetric grid. The results are shown in the diagramm. As expected, ionic strength rises as acid impurity gases dissolve into the brine. While the STIMAX=3.0 simulation (green lines) skippes geochemical calculations at stoichiometric i.s.=2.9, in both other simulation the normal ionic strength rises to nearly 6 and 8, respectively. In these cases the stoichiometric ionic strength remains at approximately 5.
Now my question is: Does TOUGHREACT check for which ionic strength is larger and compares this with STIMAX or are the presented results randomly achieved? If the latter one is true, why should the stoichiometric i.s. remain at 5, if it is allowed up to 6 / 8 (and should still rise due to the lasting supply of acid species)?
Thanks in advance!
Hi Lennard, STIMAX always applies to the true ionic strength, which is always function of concentrations of ionic species only, including primaries and secondaries. The current activity coefficient model (HKF) is valid for dilute systems and technically should not be "pushed" much above an ionic strength of ~1.5. With some NaCl dominated systems, the model can provide satisfactory results at higher ionic strength, but I would not trust computed activity coefficients for salts other than NaCl and similar salts, especially not charged S and N species.
During Newton-Raphson iterations, the true ionic strength may shoot up before converging to a reasonable value; STIMAX is used to limit these "jumps", so typically it is set to a value > 1.5 (4 or 6 is a good number). However check that the final converged results yield a (true) ionic stength < ~1.5, or be aware that at higher ionic strength the results may be significantly off. For higher salinities you will need to use the Pitzer model, which is implemented with the older V1.21 version. We have plans to implement the Pitzer model in V3 in the near future, though.