PHREEQE model

A computer code is obviously not a model. A computer code that incorporates a geochemical model is one of several possible tools for interpreting water-rock interactions in low-temperature geochemistry. The computer codes in common use and examples of their application will be the main focus of this chapter. It is unfortunate that one commonly finds, in the literature, reference to the MINTEQ model or the PHREEQE model or the EQ3/6 model when these are not models but computer codes. Some of the models used by these codes are the same so that a different code name does not necessarily mean a different model is being used. 

Uranium solubility is increased even more in the nitrate microcosms, and one possible explanation is the conversion of acetate to CO2, coupled with nitrate reduction, which would give higher dissolved carbonate concentrations in the nitrate microcosms. However, PHREEQE modelling showed that the higher CO concentration would not greatly affect uranium speciation in solution and is therefore unlikely to account for the enhanced solubility. Alternatively, as nitrate is reduced to ammonium (NH/), which promotes cation exchange, this could lead to displacement of U02 from surface complexes, which are the predominant uranyl species on mineral surfaces. " ... 

A number of models of both types have been described in the literature. Of the models, DYNAMIX would appear to have the greatest potential for use in simulating chemical transport in the deep-well environment because it incorporates the reaction-progress code PHREEQE, which can handle deep-well temperatures. PHREEQE, however, does not incorporate pressure equilibria. 

A coefficient of 0.2 is used sometimes instead of 0.3. The only variable specific to the species in question is the charge Zj, which of course is known. For this reason, the Davies equation is especially easy to apply within geochemical models designed for work at 25 °C, such as WATEQ (Ball et al., 1979) and its successors, and PHREEQE (Parkhurst et al., 1980). 

The most frequently used models are MINTEQA2 (Allison et al. 1991),WATEQ4F (Ball Nordstrom 1991), PHREEQC (PHREEQE) (Parkhurst Appelo 1999, Parkhurst 1995 Parkhurst et al. 1980) and EQ 3/6 (Wolery 1992a and 1992b). 

INTERA (1983) Geochemical models suitable for performance assessment of nuclear waste storage comparison of PHREEQE and EQ3/EQ6. INTERA Environmental Consultants, Inc., ONWI-473, 114pp. 

Plummer L. N. and Parkhurst D. L. (1990) Application of the Pitzer equations to the PHREEQE geochemical model. In Chemical Modelling of Aqueous Systems II, Symp. Ser. 416 (eds. D. C. Melchior and R. L. Bassett). American Chemical Society, Washington, DC, pp. 128-137. 

Limitations and problems additional to those of the speciation codes The models do not consider solid-solution mass transfer and provide only limited information on ion exchange/adsorption mass transfer. All the programs except PHREEQE, PHRQPITZ, and MINTEQA2 keep track of water mass. Except for EQ3/6 and the Geochemist s Workbench, rate laws for mass transfer kinetics cannot be specified. Convergence problems occur more often than for the speciation codes. 

Carbonate-rock acid-rain reaction stiochiometry is best examined using a mathematical model of the ionic species present in a dilute carbonate solution. Appropriate models have been recently developed and are used with an electronic digital computer. Although a discussion of the application of mineral-reaction modeling is outside the scope of this paper, several recent publications explain its use in the study of carbonate-mineral reactions in aqueous solution (9-11). A reaction model, such as PHREEQE (11), the model used in this study, incorporates changes in gas-solution and solution-solid equilibria, as well as ionic equilibria in solution, in arriving at a final equilibrium-solution composition (. ... 

PHREEQE A geochemical computer code based on PC (w/PHRQ- the ion-pairing model which calculates INl T and pH, redox potential and mass transfer. 

The aqueous species stability constants and individual-ion, activity-coefficient parameters from WATEQ and WATEQ2 were used as published for one set of calculations. Perchlorate and Co" were not included in the WATEQ aqueous models, but were added for these calculations. The individual-ion activity coefficient for C104 and Co were calculated using Equation 2 and ion-size parameters (Uf) of 3.5 and 6.0, respectively (16). All calculations were made using a version of PHREEQE (10) modified to calculate mean activity coefficients for salts. 

For the fit model, aqueous complexes were added as needed in order to resolve the discrepancies between the calculated and experimental mean activity coefficients. The stability constants for the complexes and the individual-ion, activity-coefficient parameters for the complexes were estimated through least-squares fitting. The program that estimated the parameters used the modified version of PHREEQE (10) as a function subroutine that calculated the mean activity coefficient The stability constants and individual-ion, activity-coefficient parameters of a specified set of complexes were adjusted until a least-squares fit was obtained between the calculated and experimental mean activity coefficients for a series of salt-solution compositions. 

Application of the Pitzer Equations to the PHREEQE Geochemical Model... 

An interactive input code for PHRQPITZ called PITZINPT (i) is analogous to the PHREEQE input code, PHRQINPT (I. The reader is referred to the PHREEQE documentation (2) and the PHRQPITZ documentation (i) for further background and modeling information. 

Many of the original convergence problems of PHREEQE were redox-related. Improvements to the convergence criteria of PHREEQE (noted in the January, 1985 version) have been incorporated in PHRQPITZ. Precautions and comments on the use of ion exchange, and titration/mixing reactions in PHREEQE, and the lack of uniqueness of modeled reaction paths (2, 39) also apply to PHRQPITZ. 

For meaningful model simulations, it is important that the thermochemical data base is accurate and internally consistent. The data base contained in PHREEQE is the same as the data base developed for WATEQ2 (30). This data base has been critically reviewed (31) and compared with WATEQFC (32). The values of Log K for solids included in this simulation, along with appropriate reactions from which they were determined, are listed in Table VIII. 

---