Mark H. Reed

Calculation of pH and mineral equilibria in hydrothermal waters with application to geothermometry and studies of boiling and dilution

Abstract Using chemical analyses and 25° pH measurements of quenched high-temperature waters, we calculate in situ pH and distribution of aqueous species at high temperature. This is accomplished by solving simultaneous mass action equations for complexes and redox equilibria and mass balance equations, on all components, including a H+ equation with as many as 60 terms (depending on water composition). This calculation provides accurate values for the activities of aqueous ions in a given water at high temperature, which are used to calculate an ion activity product (Q) for each of more than 100 minerals. The value of log(Q/K) for each mineral, where K is the equilibrium constant, provides a measure of proximity of the aqueous solution to equilibrium with the mineral. By plotting log Q/Kvs. T for natural waters, it is possible to determine: a) whether the water was in equilibrium with a host rock mineral assemblage, b) probable minerals in the equilibrium assemblage and c) the temperature of equilibrium. In cases where the fluid departs from equilibrium with a host rock assemblage, it is possible to determine whether this may result from boiling or dilution, and an estimate of amount of lost gas or diluting water can be determined. The calculation is illustrated by application to geothermal waters from Iceland, Broadlands, and Sulphur Bank, hot spring waters from Jemez, Yellowstone and Blackfoot Reservoir (Idaho) and fluid inclusions from the Sunnyside Mine, Colorado. It is shown that most geothermal waters approach equilibrium with a subsurface mineral assemblage at a temperature close to measured temperatures and that some hot springs also approach equilibrium with the host rock at temperatures above outlet temperatures but commonly below the Na-K-Ca temperatures. The log Q/K plots show that some discrepancies between Na-K-Ca temperatures on spring waters and actual temperatures result from a failure of alkali feldspars to equilibrate with the fluid and with each other. Calculations on Sulphur Bank fluids show that boiling probably caused cinnabar precipitation near 150°C and that the boiled fluids equilibrated with secondary minerals near 150° even though temperatures up to 185° have been measured at depth. For the fluid inclusions, the measured bubble temperatures are close to those calculated for equilibration of the fluid with the observed sulfide mineral assemblage. New estimates of stability constants for aluminum hydroxide complexes are included at the end of the paper.

Volatilization, transport and sublimation of metallic and non-metallic elements in high temperature gases at Merapi Volcano, Indonesia

Abstract Condensates, silica tube sublimates and incrustations were sampled from 500–800°C fumaroles and lava samples were collected at Merapi Volcano, Indonesia in Jan.–Feb., 1984. With respect to the magma, Merapi gases are enriched by factors greater than 105 in Se, Re, Bi and Cd; 104–105 in Au, Br, In, Pb and W; 103–104 in Mo, Cl, Cs, S, Sn and Ag; 102–103 in As, Zn, F and Rb; and 1–102 in Cu, K, Na, Sb, Ni, Ga, V, Fe, Mn and Li. The fumaroles are transporting more than 106 grams/day ( g d ) of S, Cl and F; 104–106 g/d of Al, Br, Zn, Fe, K and Mg; 103–104 g d of Pb, As, Mo, Mn, V, W and Sr; and less than 103 g d of Ni, Cu, Cr, Ga, Sb, Bi, Cd, Li, Co and U. With decreasing temperature (800-500°C) there were five sublimate zones found in silica tubes: 1) cristobalite and magnetite (first deposition of Si, Fe and Al); 2) K-Ca sulfate, acmite, halite, sylvite and pyrite (maximum deposition of Cl, Na, K, Si, S, Fe, Mo, Br, Al, Rb, Cs, Mn, W, P, Ca, Re, Ag, Au and Co); 3) aphthitalite (K-Na sulfate), sphalerite, galena and Cs-K. sulfate (maximum deposition of Zn, Bi, Cd, Se and In; higher deposition of Pb and Sn); 4) Pb-K chloride and Na-K-Fe sulfate (maximum deposition of Pb, Sn and Cu); and 5) Zn, Cu and K-Pb sulfates (maximum deposition of Pb, Sn, Ti, As and Sb). The incrustations surrounding the fumaroles are also chemically zoned. Bi, Cd, Pb, W, Mo, Zn, Cu, K, Na, V, Fe and Mn are concentrated most in or very close to the vent as expected with cooling, atmospheric contamination and dispersion. The highly volatile elements Br, Cl, As and Sb are transported primarily away from high temperature vents. Ba, Si, P, Al, Ca and Cr are derived from wall rock reactions. Incomplete degassing of shallow magma at 915°C is the origin of most of the elements in the Merapi volcanic gas, although it is partly contaminated by particles or wall rock reactions. The metals are transported predominantly as chloride species. As the gas cools in the fumarolic environment, it becomes saturated with sublimate phases that fractionate from the gas in the order of their equilibrium saturation temperatures. Devolatilization of a cooling batholith could transport enough acids and metals to a hydrothermal system to play a significant role in forming an ore deposit. However, sublimation from a high temperature, high velocity carrier gas is not efficient enough to form a large ore deposit. Re, Se, Cd and Bi could be used as supporting evidence for magmatic fluid transport in an ore deposit.

Contribution of C1- and F-bearing gases to the atmosphere by volcanoes

As halogen gases catalyse the destruction of ozone in the stratosphere1–3, it is important to quantify the natural emissions of halogens from active volcanoes. Here we use equilibrium thermodynamics to predict the speciation of Cl and F in volcanic gases and provide new estimates of the global emission rates to the atmosphere. Our calculations show that HCl and HF are the dominant species of Cl and F in volcanic gases, at least several orders of magnitude more abundant than all other species. We estimate the annual global volcanic fluxes of HCl and HF to be 0.4–11 Tg (1012 g) and 0.06–6 Tg, respectively. On average, <10% of these emissions come in large explosive eruptions that transmit them efficiently to the stratosphere. Although they are infrequent, large volcanic eruptions may inject significant amounts of HCl and HF into the stratosphere. Passively degassing volcanoes are also a major source of tropospheric HF, although sea salt is the main source of tropospheric HC1.

Origin, speciation, and fluxes of trace-element gases at Augustine volcano, Alaska: Insights into magma degassing and fumarolic processes

Thermochemical modeling predicts that trace elements in the Augustine gas are transported from near-surface magma as simple chloride (NaCl, KCl, FeCl2, ZnCl2, PbCl2, CuCl, SbCl3, LiCl, MnCl2, NiCl2, BiCl, SrCl2), oxychloride (MoO2Cl2), sulfide (AsS), and elemental (Cd) gas species. However, Si, Ca, Al, Mg, Ti, V, and Cr are actually more concentrated in solids, beta-quartz (SiO2), wollastonite (CaSiO3), anorthite (CaAl2Si2O8), diopside (CaMgSi2O6), sphene (CaTiSiO5), V2O3(c), and Cr2O3(c), respectively, than in their most abundant gaseous species, SiF4, CaCl2, AlF2O, MgCl2 TiCl4, VOCl3, and CrO2Cl2. These computed solids are not degassing products, but reflect contaminants in our gas condensates or possible problems with our modeling due to “missing” gas species in the thermochemical data base. Using the calculated distribution of gas species and the COSPEC SO2 fluxes, we have estimated the emission rates for many species (e.g., COS, NaCl, KCl, HBr, AsS, CuCl). Such forecasts could be useful to evaluate the effects of these trace species on atmospheric chemistry. Because of the high volatility of metal chlorides (e.g., FeCl2, NaCl, KCl, MnCl2, CuCl), the extremely HCl-rich Augustine volcanic gases are favorable for transporting metals from magma. Thermochemical modeling shows that equilibrium degassing of magma near 870°C can account for the concentrations of Fe, Na, K, Mn, Cu, Ni and part of the Mg in the gases escaping from the dome fumaroles on the 1986 lava dome. These calculations also explain why gases escaping from the lower temperature but highly oxidized moat vents on the 1976 lava dome should transport less Fe, Na, K, Mn and Ni, but more Cu; oxidation may also account for the larger concentrations of Zn and Mo in the moat gases. Nonvolatile elements (e.g., Al, Ca, Ti, Si) in the gas condensates came from eroded rock particles that dissolved in our samples or, for Si, from contamination from the silica sampling tube. Only a very small amount of rock contamination occurred (water/rock ratios between 104 and 106). Erosion is more prevalent in the pyroclastic flow fumaroles than in the summit vents, reflecting physical differences in the fumarole walls: ash vs. lava. Trace element contents of volcanic gases show enormous variability because of differences in the intensive parameters of degassing magma and variable amounts of wall rock erosion in volcanic fumaroles.

Scanning electron microscope–cathodoluminescence analysis of quartz reveals complex growth histories in veins from the Butte porphyry copper deposit, Montana

Scanning electron microscope–cathodoluminescence (SEM-CL) analysis of quartz reveals textures that cannot be observed using optical microscopy or backscattered electrons. These cryptic textures yield insight into timing and physical conditions of quartz growth, especially in environments with multiple quartz-precipitation events. Hydrothermal quartz from quartz-sulfide veins in the porphyry copper deposit in Butte, Montana, was analyzed by SEM-CL, revealing the following textures: euhedral growth zones, wide nonluminescing bands that cut across multiple quartz grains, rounded luminescent quartz grain cores with euhedral overgrowths, nonluminescing ‘‘splatters’’ of quartz connected by networks of cobweb-like nonluminescing quartz in otherwise luminescent quartz, concentric growth zones, and wide nonluminescent grain boundaries. These textures indicate that many veins have undergone fracturing, dilation, growth of quartz into fluid-filled space, quartz dissolution, and recrystallization of quartz. Precipitation and dissolution textures indicate that early quartz-molybdenite veins formed as a result of pressure fluctuations between lithostatic and hydrostatic at high temperatures, and later pyrite-quartz veins formed near hydrostatic pressure in response to temperature decrease through and beyond the field of retrograde quartz solubility.

Trace elements in hydrothermal quartz: Relationships to cathodoluminescent textures and insights into vein formation

High-resolution electron microprobe maps show the distribution of Ti, Al, Ca, K, and Fe among quartz growth zones revealed by scanning electron microscope-cathodoluminescence (SEM-CL) from 12 hydrothermal ore deposits formed between ~100 and ~750 °C. The maps clearly show the relationships between trace elements and CL intensity in quartz. Among all samples, no single trace element consistently correlates with variations in CL intensity. However in vein quartz from five porphyry-Cu (Mo-Au) deposits, CL intensity always correlates positively with Ti concentrations, suggesting that Ti is a CL activator in quartz formed at >400 °C. Ti concentrations in most rutile-bearing vein quartz from porphyry copper deposits indicate reasonable formation temperatures of <750 °C using the TitaniQ geothermometer. Titanium concentrations of <10 ppm in all veins that formed at temperatures <350 °C suggest a broad correlation between Ti concentrations and temperature of quartz precipitation. In quartz from most deposits formed at 2000 ppm, but in high-temperature quartz, Al concentrations are consistently in the range of several hundred ppm. Aluminum concentrations in quartz reflect the Al solubility in hydrothermal fluids, which is strongly dependent on pH. Aluminum concentrations in quartz therefore reflect fluctuations in pH that may drive metal-sulfide precipitation in hydrothermal systems.

As(III) and Sb(III) sulfide complexes: An evaluation of stoichiometry and stability from existing experimental data

Abstract Published experimental data on the solubility of Sb2S3 and As2S3 in sulfide solutions are used to determine the stoichiometry and stability constants of antimony and arsenic sulfide complexes at temperatures from 25 to 300°C. Least square fits and multicomponent equilibrium computations on selected solubility data indicate that the most probable stoichiometries of antimony and arsenic sulfide complexes are H2Sb2S4, HSb2S−4 and Sb2S24−, and H3As3S6, H2As3S6− and HAs3S26−. Stability constants evaluated for these complexes indicate that all three arsenic species, but only HSb2S−4, are stable within the pH range of typical geochemical systems. The determination of stability constants for the above species results from a critical evaluation of available experimental data. Many experimental studies are inconsistent with each other or, by themselves, lack sufficient quantitative information. However, when used together, the data from a few carefully selected studies yield coherent results. Newly derived stability constants for antimony hydroxide and arsenious acid indicate that, at the sulfide concentrations considered here, these species are generally unimportant relative to sulfide complexes, although they become the dominant arsenic and antimony species at high temperatures in most geothermal systems.

Theoretical Chemical Thermometry on Geothermal Waters: Problems and Methods

Using a synthetic geothermal water, we examine the effect of errors in Al analyses on theoretical chemical geothermometry based on multicomponent chemical equilibrium calculations of mineral equilibria. A new approach named FixAl that entails the construction of a modified Q/K graph eliminates problems with water analyses lacking Al or with erroneous analyses of Al. This is made possible by forcing the water to be at equilibrium with a selected Al-bearing mineral, such as microcline. In a FixAl graph, a modified Q/K value is plotted against temperature for Al-bearing minerals. Saturation indices of nonaluminous minerals are plotted in the same way as in an ordinary Q/K graph. In addition to Al concentration errors, degassing of CO2 and dilution of reservoir water interfere with computed equilibrium geothermometry. These effects can be distinguished in a Q/K graph by comparing curves for nonaluminous minerals to those of aluminous minerals then correcting for CO2 loss and dilution by a trial and error method. Example geothermal waters from China, Iceland, and the USA that are used to demonstrate the methods show that errors in Al concentrations are common, and some are severe. The FixAl approach has proved useful for chemical geothermometry for geothermal waters lacking Al analysis and for waters with an incorrect Al analysis. The equilibrium temperatures estimated by the FixAl approach agree well with quartz, chalcedony, and isotopic geothermometers. The best choice of mineral for forced equilibrium depends on pH. For most neutral pH waters, microcline and albite work well; for more acidic waters, kaolinite or illite are good choices. Measured pH plays a critical role in computed equilibria, and we find that the best pH to use is the one to which the reported carbonate also applies. Commonly this is the laboratory pH instead of field pH, but the field pH is still necessary to constrain CO2 degassing. Calculations on numerous waters in the 80–180°C reservoir temperature range indicate that mineral-aqueous equilibrium is probably nearly always achieved, but is obscured by short time-scale processes of dilution or degassing of CO2 in the near-surface environment.

Fugacity coefficients of H2, CO2, CH4, H2O and of H2O- CO2-CH4 mixtures: A virial equation treatment for moderate pressures and temperatures applicable to calculations of hydrothermal boiling

Abstract Equations derived from a quadratic virial equation in pressure, with virial coefficients expressed as a function of temperature, are fitted to published P - V - T and solubility data to yield values of second and third virial coefficients for pure and mixed gases. These coefficients are not virial coefficients sensu stricto and are used to compute fugacity coefficients of pure H 2 , H 2 O, CO 2 and CH 4 , and of mixed H 2 O, CO 2 and CH 4 , and to estimate enthalpies for these gases. For H 2 , the P-T range of application is from 25 to 600 C and up to 3000 bars, and for CH 4 , from 16 to 350 C and up to 500 bais. For H 2 O and CO 2 , two P-T ranges are considered: below 350 C, up to 500 bars, and from 450 to 1000 C, up to 1000 bars. The method presented here is limited to the P-T range of the fitted experimental data, and cannot represent accurately P-V-T data close to the critical region. This virial equation treatment yields simple analytical expressions that are suitable for multicomponent equilibrium calculations. Examples of equilibrium calculations between aqueous and gas phases show that ideal mixing of real gases is a sufficient approximation for modeling boiling in geothermal and epithermal systems. However, non-ideal mixing has to be considered for aqueous-gas systems at pressures much higher than the saturation pressure of pure water.

Reconstruction of in situ composition of sedimentary formation waters

Abstract Chemical equilibrium calculations on sedimentary formation waters show that the waters, as analyzed, cannot be in equilibrium with diagenetic minerals in their host rocks at the formation temperature. However, if alkalinity is corrected to account for organic acid anions, and if the pH and bicarbonate are corrected for CO 2 loss from the sample, chemical equilibrium between formation waters and host rock diagenetic minerals can be clearly shown for systems in the temperature range of 75 to 160°C. Compositional reconstruction of some formation waters from published analyses is complicated by lack of analytical data for aluminum, silica, and organic acid anions. Missing aluminum and silica can be estimated by assuming equilibrium with an aluminum silicate (K-feldspar, muscovite) and quartz or chalcedony. pH, CO 2 , and organic acid anions can be reconstructed by fixing CO 2 to exactly saturate calcite at the formation temperature because the fast kinetics of calcite precipitation makes it almost certain that calcite saturation is more likely than the strong supersaturation that is otherwise observed. Results from the equilibrium calculations are evaluated by using graphs of the saturation states of diagenetic minerals vs. temperature, for each of many sedimentary brines. If the diagenetic minerals selected as diagnostic of equilibrium (from qz, chalcedony, mus, paragonite, k-sp, alb, kaol, ca, and dol) are not saturated at or near a single temperature, the missing or erroneous quantities of components are adjusted to obtain agreement in the saturation temperature. Composition data for fluids from four locations are used in the calculations: Kettleman North Dome, California, offshore Norway, the Texas Gulf Coast, and offshore Texas. The calculations suggest that in most cases, control of silica concentration shifts from chalcedony to quartz with increasing temperature near 100°C. In some fluids, silica concentration may approach chalcedony saturation to temperatures exceeding 150°C, where there are shale units containing smectite undergoing the smectite to illite reaction. Deviation in silica activity from equilibrium with chalcedony or quartz is small for most of the fluids, and may result from precipitation of silica as polymers or amorphous solids upon cooling, and either removal of precipitates upon filtering before analysis, or nonreactivity of the precipitates in the analytical method used. Four fluids containing significant iron and having apparently degassed significant CO 2 also show substantial apparent silica loss, and therefore, silica loss most probably results from the precipitation of amorphous Fe-silicate caused by pH increase due to degassing, and by cooling. The methods used here can be applied as a geothermometer to predict formation temperatures, and, when applied to Kettleman North Dome, yield a thermal gradient of 37.1°C/km. Formation temperature data for the Texas waters are in agreement with equilibrium temperatures predicted by the calculations.