Thanks for the additional information.
However, I'm still not entirely clear how to proceed with this.
For example, if I want to model exchange on smectite, which say has 4 exchange components (e.g. Na, K, Ca, Mg), do I need to have 4 versions of the IonEx database for each of the components, i.e. one each for Na, K, Ca and Mg?
It would be a lot easier to do this if the input requirements for GWB were like those in PHREEQC - presumably this would be easy to implement? Also, in PHREEQC you can calculate the exchanger composition in equilibrium with a given water composition, which is also a very useful feature.
Whilst I'm on the subject of GWB developments, I'm really surprised that there's no capability for modelling solid-solutions. Again, this capability is in PHREEQC, and as I recall, was available in EQ3/6 (albeit for ideal solid solutions) about 20 years ago. Is this something that you can include for the next version?
Many thanks for your help.