Mineral solubilities in concentrated brines

Weare, J.H., 1987, Models of mineral solubility in concentrated brines with application to field observations. In I. S.E. Carmichael and H.P Eugster (eds.), Thermodynamic Modeling of Geological Materials Minerals, Fluids and Melts. Reviews in Mineralogy 17,143-176. 

To calculate gas solubility in natural geochemical systems, basic thermodynamic properties such as the Henry s law constant and, in the case of weak electrolytes the dissociation constant, must be combined with a thermodynamic model of aqueous solution behavior. An analogous approach has been used to predict mineral solubilities in concentrated brines (1). Such systems are also relevant to the atmosphere where very concentrated solutions occur as micrometer sized aerosol particles and droplets, which contain very small amounts of water relative to the surrounding gas phase. The ambient relative humidity (RH) controls solute concentrations in the droplets, which will be very dilute near 1(X)% RH, but become supersaturated with respect to soluble constituents (such as NaCl) below about 75% RH. The chemistry of the aerosol is complicated by the non-ideality inherent in concentrated electrolyte solutions. 

Virial methods, the second type, employ coefficients that account for interactions among the individual components (rather than species) in solution. The virial methods are less general, rather complicated to apply, require considerable amounts of data, and allow little insight into the distribution of species in solution. They can, however, reliably predict mineral solubilities even in concentrated brines. 

DH-type, low ionic-strength term. Because the DH-type term lacks an ion size parameter, the Pitzer model is also less accurate than the extended DH equation in dilute solutions. However, a.ssuming the necessary interaction parameters (virial coefficients) have been measured in concentrated salt solutions, the model can accurately model ion activity coefficients and thus mineral solubilities in the most concentrated of brines. 

Freeze concentration processes are based on the difference in component concentrations between solid and liquid phases that are in equihbrium. Most minerals and many organics grow less soluble in water as the temperature decreases. When an aqueous solution is cooled, ice usually crystallizes as a pure material, and dissolved components in the aqueous waste stream are concentrated in the remaining brine, thereby reducing the volume of waste. 

In spite of these major limitations, considerable progress has been made in understanding many of the important factors that influence subsurface water chemistry. Na+, Ca2+ and Cl- account for the major portion of dissolved components in most brines (Figure 8.6). Ca2+, which can comprise up to 40% of the cations, usually increases relative to Na+ with depth (Figure 8.7). Br and organic acids are commonly found at concentrations of 1 to 2 g L1 (Land, 1987). The bicarbonate concentration is largely limited by carbonate mineral solubility, and sulfate is generally found in low concentrations as a result of bacterial and thermal reduction processes. 

Land (1987) has reviewed and discussed theories for the formation of saline brines in sedimentary basins. We will summarize his major relevant conclusions here. He points out that theories for deriving most brines from connate seawater, by processes such as shale membrane filtration, or connate evaporitic brines are usually inadequate to explain their composition, volume and distribution, and that most brines must be related, at least in part, to the interaction of subsurface waters with evaporite beds (primarily halite). The commonly observed increase in dissolved solids with depth is probably largely the result of simple "thermo-haline" circulation and density stratification. Also many basins have basal sequences of evaporites in them. Cation concentrations are largely controlled by mineral solubilities, with carbonate and feldspar minerals dominating so that Ca2+ must exceed Mg2+, and Na+ must exceed K+ (Figures 8.8 and 8.9). Land (1987) hypothesizes that in deep basins devolatilization reactions associated with basement metamorphism may also provide an important source of dissolved components. 

Experimental investigations of carbon dioxide mineral trapping have been started, but there is little understanding of the processes involved on the molecular level, due to the complex nature of the brine and the physical conditions present in the brine aquifers. One such complexity issue is the dissolution of CO2 from the gaseous phase into aqueous solution. This process is thermodynamically unfavorable with a ArG° value of 2.00 kcal/mol, in pure water at STP [3]. This becomes even more thermodynamically unfavorable as salts are introduced into the solution. Figure 17.1 shows how the solubility of CO2 in aqueous solution is also dependent upon the salt concentration and salt composition, even from simple salt solutions. According to Fig. 17.1, there is no correlation with size and the ability for the solution to uptake the C02. 

For the three MVT models, total chloride concentrations are near 3 molal, whereas in the RBRBM model total chloride is 2 molal. For the four ore-fluid models, several mineral saturation constraints were applied. These are listed in Table 6 together with the concentrations of organic ligands used in these models. The concentrations of acetate, oxalate, malonate, and succinate are based on maximum values reported for basinal brines (Table 2). The concentration of total catechol (0.01 molal) is an arbitrarily chosen value that is in the concentration range reported for malonate, but is well below catechol solubility in water (4 molal at 25 °C). The total chloride to total acetate ratio is about 30 1 for the three MVT models and 20 1 for the RBRBM composite fluid. 

The resulting Soln. C is a predominantly NaCl solution similar to terrestrial seawater (Soln. D, Table 5.3). Had we chosen a concentration factor of 600-fold, the agreement would have been even better. In any case, the concentration factor is arbitrary. The point is that simple processes, starting with a dilute Fe-Mg-HC03-rich solution formed by reaction of water with ultra-mafic and mafic rocks, evaporation, and carbonate precipitation, converted the solution into an Earth-like seawater NaCl brine. The Na/Mg ratio of solution C is 9.9, while terrestrial seawater has a Na/Mg ratio of 8.8 (Soln. 5.3D). Note also the similar pH values (8.03 and 8.05, Table 5.3). This solution did not (cannot) evolve into an alkali soda-lake composition as some have hypothesized or assumed for Mars (e.g., Kempe and Kazmierczak 1997 Morse and Marion 1999) because the mass of hypothesized soluble iron and magnesium and the low solubility of their respective carbonate minerals are sufficient to precipitate most of the initial soluble bicarbonate/carbonate ions. 

Chloride minerals are rarely found in coal in the form of solid species because of high solubility of sodium, calcium and trace metal chlorides in coal strata waters. The "inherent" water content of coal is related to its porosity and thus the moisture content of lignite deposits can exceed 40 per cent decreasing to below 5 per cent in fully bituminous coals (11). Chlorides, chiefly associated with sodium and calcium constitute the bulk of water-soluble matter in British bituminous coals (12). Skipsey (13) has found that the distribution of chlorine coals was closely related to the salinity of mine waters. Hypersaline brines with concentrations of dissolved solids up to 200 kg m occur in several of the British Coalfields. 

---