Activity calculation model

[GeeqC]

Dear Dr. Parkhurst,

I am having some problems understanding which model is used for activity correction. So far I assumed that the activity model is automatically selected according to the database we choose. For example, the Pitzer model is used if we select the pitzer.dat database or the SIT model is used if we select the sit.dat database. However, I noticed that also e.g. the phreeqc.dat database contains pitzer-parameters.

My question is, which models are used: Debye-Hückel, extended Debye-Hückel, SIT after Guggenheim et al. (1955), SIT after Helgeson et al (1981) or Pitzer? Where can we find the actual equations?

Thank you very much!

GeeqC

[Pak]

That is explained in the user Manual. In the block related to SOLUTION_SPECIES (Description of Data Input page 201). Is the -gamma parameter, at least for Debye-Hückel and Davies. For using SIT and Pitzer not sure, but it should be probably in the manual.

[dlparkhurst]

If the keyword data block PITZER is in the database (pitzer.dat, and others), then the Pitzer model is used. If the keyword data block SIT is in the database (sit.dat), then the SIT model is used. If the keyword data block LLNL_AQUEOUS_MODEL_PARAMETERS (llnl.dat), then the Lawrence Livermore National Labs version of Debye Huckel is used. For all other databases, the Davies or Wateq Debye Huckel equations are used, where definitions in SOLUTION_SPECIES determine which equation is used.

phreeqc.dat contains no Pitzer parameters.

Documentation for Pitzer equations are in the PHRQPITZ manual (https://pubs.er.usgs.gov/publication/wri884153) as well as the papers by Pitzer and coworkers, and other authors. SIT (Specific ion Interaction Theory) equations can be found in papers by Grenthe. Documentation of Debye-Huckel equations is in the manual for PHREEQC version 2 (https://pubs.er.usgs.gov/publication/wri994259), which is also included in the docs directory of all program distributions. Additional information and citations on pressure dependence can be found in the notes at the end of phreeqc.dat.

[GeeqC]

Dear Dr. Parkhurst,

I just saw that you wrote a further response. This fully clarifies my question.

Thanks again!
GeeqC

[d_guse]

Dear Mr. Parkhurst,

I have two (late) follow-up questions.

1. In the documentation it is stated that either the SIT data block or the PITZER data block can be used in one Phreeqc run which also fits with your statement here. Accordingly, there is an error message if someone still tries to use both. However, there is no error message when using both LLNL_AQUEOUS_MODEL_PARAMETERS and PITZER or LLNL_AQUEOUS_MODEL_PARAMETERS and SIT data blocks. So it is possbile to combine both or which method is actually used in the background?

2. You also wrote that, e.g., if the PITZER data block is used, the Pitzer model is used.
a) What happens if no Pitzer coefficients are avaible for some species? Is Davies then used instead? Or are the eq. (2), (3) and (7) for gamma_cation, gamma_anion and F given in the documentation for PHRQPITZ by Plummer et al. 1988 simplified to ln(gamma) = z^2 * F with F = f(A, b, I) since all other parameters (beta_i, C_i, Phi_i, etc.) are zero?
b) What happens if -gamma or -llnl_gamma values are specified for some species (while a PITZER data block is used for other species)? Are they ignored or is the WATEQ/ LLNL equation used for those species instead?
c) What happens if -gamma values as well as Pitzer coefficients are availabe for a species? Is the Pitzer model then prioritized?

I hope that these questions are not too trivial. Thank you very much.


Kind regards
David

[dlparkhurst]

I think if a SIT data block is included, the calculations will be SIT and  LLNL_AQUEOUS_MODEL_PARAMETERS definitions will have no effect. Same with PITZER. However, to be sure, define at most one: SIT, PITZER, or LLNL_AQUEOUS_MODEL_PARAMETERS.

The values of any interaction parameters that are not defined are effectively zero. There is a D-H term in the Pitzer equations; no other extended D-H (like Davies) is used if parameters are zero.

Only Pitzer parameters (C, beta, etc) are used in calculating activity coefficients when a PITZER data block is defined.

In short, you have to use only one of the following approaches, (1) ion-association with activity coefficient parameters a and b (this includes the llnl approach), (2) PITZER, or (3) SIT. There are no options for hybrid calculations.