Pytzer: the Pitzer model for chemical activities and equilibria in aqueous solutions in Python

Changes in v0.5.3

• Created new Humphreys22 parameter library and validated its calculations against HWT22.
• Created new Clegg22 parameter library (but its calculations against CHW22 still need validating).
• Created new Clegg23 parameter library and validated its calculations against CWTD23.
• Renamed io module as get.

Pytzer: the Pitzer model for chemical activities and equilibria in aqueous solutions in Python

Changes in v0.5.2

• Fixed calculation error in convert.solvent_to_solution.
• Added more ions in properties.ion_to_mass.

Pytzer: the Pitzer model for chemical activities and equilibria in aqueous solutions in Python

Changes in v0.5.1

• Fixed pz.convert.solvent_to_solution to take and return pK values instead of K values, and to replace any NaNs in the input with 0.
• Added pz.equilibrate.stoichiometric.get_constants() to calculate stoichiometric equilibrium constants for the carbonic acid equilibria when total dissolved inorganic carbon is zero.

Pytzer: the Pitzer model for chemical activities and equilibria in aqueous solutions in Python

Changes in v0.5.0

• Switched from Autograd to JAX for faster automatic differentiation and JIT compilation.
• Renamed and reorganised virtually everything.

Demonstration preparation

• Added osm function for osmotic coefficient to matrix module;
• Added io module functions to roughly convert between molinity (mol/kg-solution) and molality (mol/kg-H₂O);
• Added salinity2mols function in io module to estimate molalities from salinity for MZF93 artificial seawater;
• Switched assemble function in matrix module to use the MarChemSpec parameter library by default.

Basic plotting functions

• Removed pressure input from thermodynamic equilibrium constant functions;
• Converted copies from shallow to deep in parameter library modules;
• Added plot module with some meta-analytical visualisation tools (not yet documented).

Now with carbonate and borate equilibria

• Added carbonic acid and boric acid equilibrium solving;
• Eliminated module tables;
• Updated functions and parameter libraries noting that ['Cl', 'CO3'] gets sorted to ['CO3', 'Cl'];
• Fixed incorrect alph1 values for all HMW 1:(>1) electrolytes;
• Added several new interaction parameter functions;
• Testing MIAMI parameter library against PM16's model:
  • Adjusted numerical values to fix errors in some parameter functions:
    • bC_Ca_BOH4_SRM87;
    • bC_K_OH_MP98;
    • bC_Mg_BOH4_SRM87;
    • bC_Na_Br_MP98;
    • bC_Na_F_MP98;
    • bC_Na_HSO4_MP98;
    • bC_Sr_Cl_MP98;
    • theta_K_Sr_MP98;
  • Added new MP98 functions until numerical conflicts can be resolved:
    • bC_K_CO3_MP98 should follow SRG87;
    • theta_H_K_MP98 and theta_H_Na_MP98 should follow CMR93.

Equilibrium

• Module dissociation added with functions to evaluate thermodynamic dissociation constants;
• Module equilibrate added with an equilibrium solver and associated functions;
• Added blackbox_equilibrate function to demonstrate equilibration procedure;
• Fixed indexing error in matrix.assemble for the zeta coefficient matrix;
• Module jfuncs renamed as unsymmetrical;
• As a general terminology change, the Pitzer model interaction coefficients are now instead referred to as parameters, for better consistency with the published literature. As a consequence:
  • Module coefficients renamed as parameters;
  • Module cflibs renamed as libraries;
  • Class CoefficientLibrary renamed as ParameterLibrary, and its abbreviation cflib to prmlib;
• Parameter libraries can now be stored in separate script files.
• Added complete HMW84 parameter library;
• Added all interaction coefficients to MIAMI parameter library - but its calculation results have not yet been tested;
• Fixed multiplier error for Cphi values in PM73_TableVI in tables module;
• Added presumably missing *1e-3 multiplier to Cphi values in all SRRJ87 binary interaction functions;
• Added meta-analysis functions in meta module.

Matrix madness

• Module props renamed as properties;
• Module coeffs renamed as coefficients;
• Class CoeffLib renamed as CoefficientLibrary;
• Added docstrings to all functions in coefficients module;
• Added matrix module with alternative, matrix-based Pitzer model implementation;
• Added none function to jfuncs module to ignore unsymmetrical mixing;
• Fixed A_ϕ bug in CoefficientLibrary method print_coeffs;
• In preparation for equilibration calculations:
  • Added gettots function to io module.

Under pressure

• Moved Debye-Hückel functions (for A_ϕ) from coeffs into new module debyehueckel;
• Added Archer and Wang (1990) calculation of A_ϕ;
• Added pressure input (pres in dbar) to all relevant functions in coeffs, debyehueckel, model, io and blackbox functions;
• Added teos10 module to calculate various properties of pure water;
• Added Seawater coefficient library - like MarChemSpec, but with pressure-dependent A_ϕ term (following Archer and Wang, 1990), and other general improvements;
• Made cflib input optional for all model functions (default = Seawater);
• Fixed loop index correction for n-n' interactions;
• Eliminated unnecessary arrays for constant values in coeffs;
• Adjusted model.B and model.CT function inputs to reduce number of cflib function calls;
• Rewrote CRP94 Aosm function to make it autograd-able;
• Added docstrings throughout (except for functions in coeffs).