LLNL Database

For a number of reasons, using saturation indices as measures of the mineral masses to be formed as a fluid approaches equilibrium is a futile (if commonly undertaken) exercise. First, a mineral s saturation index depends on the choice of its formula unit. If we were to write the formula for quartz as Si2C>4 instead of Si02, we would double its saturation index. Large formula units have been chosen for many of the clay and zeolite minerals listed in the llnl database, and this explains why these minerals appear frequently at the top of the supersaturation list. 

A flexible method for modeling redox disequilibrium is to divide the reaction database into two parts. The first part contains reactions between the basis species (e.g., Table 6.1) and a number of redox species, which represent the basis species in alternative oxidation states. For example, redox species Fe+++ forms a redox pair with basis species Fe++, and HS- forms a redox pair with SO4. These coupling reactions are balanced in terms of an electron donor or acceptor, such as 02(aq). Table 7.1 shows coupling reactions from the llnl database. 

Table 7.1. Some of the redox couples in the LLNL database...

Fig. 8.2. Values of A, B, and B for the B-dot (modified Debye-Huckel) equation at 0 °C, 25 °C, 60 °C, 100 °C, 150 °C, 200 °C, 250 °C, and 300 °C (squares) and interpolation functions (lines). Values correspond to I taken in molal and a in A. Data from the LLNL database, after Helgeson (1969) and Helgeson and Kirkham (1974).

As an example of an equilibrium calculation accounting for surface complexation, we consider the sorption of mercury, lead, and sulfate onto hydrous ferric oxide at pH 4 and 8. We use ferric hydroxide [Fe(OH)3] precipitate from the LLNL database to represent in the calculation hydrous ferric oxide (FeOOH /1H2O). Following Dzombak and Morel (1990), we assume a sorbing surface area of 600 m2 g-1 and site densities for the weakly and strongly binding sites, respectively, of 0.2 and 0.005 mol (mol FeOOH)-1. We choose a system containing 1 kg of solvent water (the default) in contact with 1 g of ferric hydroxide. 

Two methods of balancing reactions are of interest. We can balance reactions in terms of the stoichiometries of the species considered. In this case, the existing basis B is a list of elements and, if charged species are involved, the electron e. Alternatively, we may use a dataset of balanced reactions, such as the llnl database. Basis B, in this case, is the one used in the database to write reactions. We will consider each possibility in turn. 

Substituting values of AG/ taken (in kJ mol 1) from Robie el al. (1979) and the LLNL database, the equation,... 

A reaction dataset, such as the LLNL database, provides an alternative method for balancing reactions. Such a database contains reactions to form a number of aqueous species, minerals, and gases, together with the corresponding equilibrium constants. Reactions are written in terms of a generic basis set B, which probably does not correspond to set B, our choice of species to appear in the reaction. 

The inverted matrix is the transformation matrix for the basis we have chosen. The reaction in the LLNL database for Ca-clinoptilolite,... 

To further illustrate how the basis-swapping algorithm can be used to balance reactions, we consider several ways to represent the dissolution reaction of pyrite, FeS2. Using the program RXN, we retrieve the reaction for pyrite as written in the llnl database... 

For the Hveragerdi 4 well, we follow the same procedure, using the data in Table 23.2 and the calculations already shown. In this case, the model predicts that a number of minerals in the LLNL database are supersaturated near the inflow temperature of 181 °C. Close examination reveals that each of the supersaturated minerals contains either Mg++, Ca++, or Fe++, components that are characteristically depleted in geothermal fluids. The Mg++ concentration in this fluid, for example, is just 2 p,g kg-1. 

For the Hveragerdi 4 well, we follow the same procedure, using the data in Table 17.2 and the calculations already shown. In this case, the model predicts that a number of minerals in the llnl database are supersaturated near the inflow... 

Figure 10.5 Solubility of amorphous silica as calculated using PHREEQC with K values from the LLNL database.

---