Tf. Xu et al., Modeling of non-isothermal multi-component reactive transport in field scale porous media flow systems, J HYDROL, 214(1-4), 1999, pp. 144-164
A general 2-D finite element multi-component reactive transport code, TRANQ
UI, was developed, using a sequential iteration approach (SIA). It is well
suited to deal with complex real-world thermo-hydro-geochemical problems fo
r single-phase variably water saturated porous media flow systems. The mode
l considers a wide range of hydrological and thermodynamic as well as chemi
cal processes such as aqueous complexation, acid-base, redox, mineral disso
lution/precipitation, gas dissolution/ex-solution, ion exchange and adsorpt
ion via surface complexation. Under unsaturated conditions only water flow
is considered, although gas pressures are allowed to vary in space in a dep
th-dependent manner specified by the user. In addition to the fully iterati
ve sequential approach (SIA), a sequential non-iterative approach (SNIA), i
n which transport and chemistry are de-coupled, was implemented and tested.
The accuracy and numerical performance of SIA and SNIA have been compared
using several test cases. The accuracy of SNIA depends on space and time di
scretization as well as on the nature of the chemical reactions. The capabi
lity of the code to model a real case study in the field is illustrated by
its application to the modeling of the hydrochemical evolution of the Llobr
egat Delta aquitard in northeastern Spain over the last 3500 years during w
hen fresh-water flow from a lower aquifer displaced the native saline aquit
ard waters. Manzano and Custodio carried out a reactive transport model of
this case study by using the PHREEQM code and considering: water flow, aque
ous complexation, cation exchange and calcite dissolution. Their results co
mpare favorably well with measured porewater chemical data, except for some
of the cations. Our code is not only able to reproduce the results of prev
ious numerical models, but leads to computed concentrations which are close
r to measured data mainly because our model takes into consideration redox
processes in addition to the processes mentioned above. A number of sensiti
vity runs were performed with TRANQUI in order to analyze the effect of err
ors and uncertainties on cation selectivities. (C) 1999 Elsevier Science B.
V. All rights reserved.