Steven J. Fritz

Bacterial production of methane and its influence on ground-water chemistry in east-central Texas aquifers

Geochemical and isotopic data for methane and ground water indicate that gaseous hydrocarbons in Eocene aquifers in east-central Texas form by bacterial processes. The δ13C values of methane from five wells in the clay-rich Yegua and Cook Mountain Formations range from -71‰ to -62‰. Methane from ten wells in the cleaner sands of the Sparta and Queen City Formations have δ13C values between -57‰ and -53‰. The carbon isotopic difference between methanes from the Yegua and Sparta aquifers is comparable to the isotopic difference in sedimentary organic matter from outcrops of the units, suggesting substrate control on the δ13C of bacterial methane. Hydrogen isotopic compositions of methane from the aquifers are similar, averaging -181‰. This high value suggests methane production predominantly by CO2 reduction. The δ13C of dissolved inorganic carbon (DIC) in high bicarbonate waters increases from about -20‰ to 0‰ with increasing DIC. Mass-balance calculations indicate that the DIC added to the ground water has δ;13C values as high as 10‰. This 13C-enriched carbon is predominantly derived from CO2 production by fermentation and anaerobic oxidation reactions combined with CO2 consumption by CO2 reduction. This process is responsible for high bicarbonate contents in these and probably other Gulf Coast ground waters.

Electrolyte-induced solute permeability effects in compacted smectite membranes

Abstract Experiments suggest that increasing electrolyte flux into a compacted smectite membrane may result in greater permeability of the membrane to the electrolyte. The influx of electrolyte into the membrane pores decreases the double layer thickness, therefore reducing the clays ability to exclude anions. As a result, solute previously accumulated in the concentration polarization layer (CPL) may be able to dissipate through the membrane, producing a temporary situation where the effluent is more concentrated than the feed solution. Of course, at some juncture the CPL will diminish sufficiently so that the effluent concentration will be the same as that of the feed solution and a true state will be reached. The fact that increasing electrolyte concentration within the clay membrane decreases the membrane efficiency implies that the result of solute accumulation in geological basins, even though a portion of such accumulation is attributable to the solute filtering ability of low permeability rocks, is the ultimate cessation of membrane effects. At the start of the cycle, shale membranes have low solute concentrations and thus high rejection capabilities. The largest amount of solute accumulates deep in the basin center, where vertical flux rates are high. Because of greater solute concentration within the aquifer, the solute concentration in the centrally located membranes rises more quickly than that of those located near the basin margins. Therefore, the cessation of membrane effects occurs first in the deep center of the basin and gradually moves upward and outward toward the basin margins. The fact that membrane efficiency decreases as pore solute concentration increases might help explain why conclusive trends attributable to membrane effects have not been more commonly observed in the subsurface.

Characterizing shallow aquifers using tritium and 14C: periodic sampling based on tritium half-life

Abstract Thirteen water-production wells in glacial sediments in and around Waterloo, Ontario, (Canada) were sampled for major ions, 3H, 13C, 14C and 18O in 1976 and again in 1988 to gauge the movement of bomb-spike tritium and 14C through the Waterloo aquifer and also to assess the reliability of a hydrogeochemical interpretation of the dynamic character of an aquifer based only on single sampling episode. The suite of samples collected in 1976 was not able to indicate how fast the tritium spike was moving or even if each well-screen was intercepting the front end or tail end of the bomb-spike peak corresponding to infiltration of mid-1960s precipitation. The second sampling event portrayed movement which, qualitatively at least, gave information relating to the more permeable zones of the aquifer and direction of water movement. Water from all but one of the 13 wells sampled in 1988 registered a drop in tritium relative to the 1976 sampling. A well is deemed to be intercepting the front end of the peak of the bomb-spike if its 1988 tritium activity (±2 TU analytical uncertainty) is greater than one half that of the tritium activity of well water measured in 1976 (±8 TU uncertainty). Only one well and two shallow piezometers met this criterion. Ten wells had TU1988/TU1976 ratios which could be interpreted as either greater than or less than 0.5, if the analytical uncertainties of both analyses in this ratio are rigorously applied. Screens of two wells intercept the tail end of the spike because both their uncertainty-adjusted TU1988/TU1976 ratios range from 0.29 to 0.41. Carbon-14 activity for individual wells varied by no more than 6 Per cent Modern Carbon (PMC) between 1976 and 1988. Bomb-spike 14C is not so easily detected as tritium in passage through aquifers because the thermonuclear input of 14C into the atmosphere was much less intense (relative to pre-bomb, background levels) than that for tritium. Also, incongruent dissolution of dolomite, coupled with differing dissolution kinetics between dolomite and calcite, precipitates 14C-bearing calcite in the saturated zone. Although the chemistry of an aquifer can be reasonably characterized by a single sampling episode, the recharge rate and groundwater flow paths are best delineated from a geochemical perspective by multiple samplings in which a persistent chemical and/or isotopic tracer is sought out and repeatedly analyzed. In the Waterloo aquifer, these are tritium and Cl. The sameness of 14C activity in waters sampled 12 a apart makes this a poor candidate to use in this scheme.

A comparative study of gabbro and granite weathering

In the piedmont of the southeastern U.S.A., gabbros and granites weather in quite different fashions. Granites develop extensive weathering-rind systems (> 2 m thick) in which plagioclase is altered to a gibbsite-kaolinite assemblage in rindlets adjacent to the fresh-rock contact. In the outer rindlets, the clay fraction is dominated by kaolinite and dioctahedral vermiculite. Not surprisingly, granitic weathering-rind systems display relative depletions of Si, Na, Ca, Rb and Sr and relative enrichments of sesquioxides. In contrast, gabbro weathering-rind systems are characterized by many thin (< 1 cm) rindlets whose alteration products are dominated by limonitization of ferromagnesian minerals. Apart from a decrease in FeOFe2O3 ratios in the outer rindlets, there is little relative depletion or enrichment of major elements in gabbro weathering rinds as compared to their fresh-rock cores. This contradiction in Goldichs stability series is explained if weathering rinds are produced faster on gabbros than in granite outcrops. Thus the low degree of mineral alteration in gabbro rindlets reflects a shorter time of residence for minerals in this system compared to rindlets on granitic outcrops. Biotite and perthitic microcline were separated from a fresh granite and its attendant weathering-rind system. During alteration, the albitic blebs of the perthite are dissolved, leaving no trace of residual clay. The result is a porous microcline whose composition is equivalent to the microcline matrix in the unweathered perthite. Progressive alteration of biotite results in oxidation of octahedrally-coordinated Fe2+ with concomitant loss of K. This results in the formation of dioctahedral vermiculites in outer rindlet samples of granite.

Measuring the ratio of aqueous diffusion coefficients between 6Li+Cl− and 7Li+Cr− by osmometry

Abstract Osmotic equilibrium is a singular occurrence in the evolution of an osmotic cell because at this event the net solution flux is zero such that −J w · V w = J s · V s . At this juncture, the diffusion coefficient of the solute through the membrane (ω) equals the solute flux (Js) divided by the osmotic pressure (ΔΠ). Because the solute permeability coefficient (ω) is related to the Fickian diffusion coefficient (D) through the gas constant, temperature, and the membranes thickness and tortuosity, the ratio of ω values for individual isotopic species equals the ratio of D values for the same isotopic components. A 0.9450 molal LiCl solution was placed within sealed dialysis tubing and osmoted against a kilogram of deionized water at 22°C. Osmotic equilibrium occurred at 164 ± 10 min. The ratio of ω6 Li + Cl − ω7 Li + Cl − was measured to be 1.011 ± 0.003—a value close to the square root of the mass ratio between 7LiCl and 6LiCl (= 1.012) as calculated by Grahams Law. The measured diffusion coefficient ratio was used to predict the degree of hyperfiltration-induced fractionation of Li isotopes as a function of membrane ideality. When a membranes σ exceeds 0.95 (as is likely for low-porosity shales) the 6 Li 7 Li ratio on the high-pressure side of the membrane can theoretically vary by more than 0.0017.

Determination of Fe3+/Fe using the electron microprobe: A calibration for amphiboles

Abstract Iron is a common constituent in minerals from the Earth’s crust and upper mantle and often occurs in minerals as mixtures of two valence states, Fe3+ or Fe2+. Quantification of the values of Fe3+/FeTotal, where FeTotal = Fe3++Fe2+, in minerals may be necessary to accurately apply certain mineral equilibria to determine equilibrium values of important variables such as temperature (T), pressure (P), and oxygen fugacity (ƒO₂). Most useful would be an analytical technique that permits determination of values of Fe3+/FeTotal within a single mineral grain that is contained within a standard petrographic thin section, and the excellent spatial resolution and relative accessibility of the electron microprobe (EMP) have resulted in various attempts to use this instrument to determine values of Fe3+/FeTotal. These efforts have typically involved quantifying characteristics of the FeLα and/or FeLβ peaks. In this paper, we employ the method of Fialin et al. (2001), who have shown that the location of the FeLα peak changes as a function of Fe content and values of Fe3+/ FeTotal, to determine values of Fe3+/FeTotal in amphiboles. We have characterized the FeLα peak in several amphiboles with known values of Fe3+/FeTotal using the electron microprobe at Texas A&M University. Initial analyses employed a beam current of 20 nA in an effort to avoid Fe-oxidation due to electron beam generated H-loss (Wagner et al. 2008). Subsequent analyses were conducted at 100 nA, and the results are consistent with the 20 nA data only when relatively short duration analytical times were used. The position of the FeLα peak was determined for three suites of amphiboles that have been experimentally treated such that grains in any one of these mineral suites are chemically identical except for differences in the values of Fe3+/FeTotal. A linear relation between the FeLα peak location and value of Fe3+/FeTotal was observed for each of these three amphibole suites. These three lines differ from one another in both their slope and intercept and these differences vary as a function of Fe content. Thus, these amphiboles served as the basis for the derivation of a relation between Fe content and FeLα peak location, both measured with the EMP, and the value of Fe3+/FeTotal as originally determined with 57Fe Mössbauer spectroscopy. The relation between the relative peak position (RPP = hematite standard FeLα peak position - amphibole FeLα peak position), Fe content, and Fe3+/FeTotal is Fe3+/FeTotal = RPP - RPP(0)/RPP(1) - RPP(0), where RPP(0) = -1.37 × FeO2 + 19.59 × FeO - 3.85, RPP(1) = -1.25 × FeO2 + 21.39 × FeO + 13.05, and FeO refers to the wt%FeO. This relation reproduces the measured values of Fe3+/FeTotal to within ±0.07 and, therefore, should permit determination values of Fe3+/FeTotal in amphiboles with Fe contents from 7 to 13 wt% FeO with similar precision. The amphiboles that were used in this study were kaersutites, Ti-bearing pargasites, and pargasitic hornblendes. The calibration presented here should, at the very least, be applicable to amphiboles with similar compositions, and although further verification is necessary, this calibration may be useful for determining values of Fe3+/FeTotal in amphiboles with distinctly different compositions and may even be more universally applicable.

Isotopic fractionation and overall permeation of lithium by a thin-film composite polyamide reverse osmosis membrane

Abstract A small, but significant fractionation of lithium isotopes was observed during reverse osmosis experiments performed with a thin-film composite polyamide membrane (FilmTec FT-30 membrane). Relative to the feed solution, the heavier lithium isotope 7 Li was depleted in the permeate from 4.3 to 10.6 per mil (±3.3 per mil) during six experimental runs. The observed isotopic fractionation might be the result of the slightly different permeabilities of 6 Li and 7 Li across the FT-30 membrane. The heavier of two solute isotopes has a slightly lower mobility due to its slightly greater mass. Therefore, the heavy isotope permeates slightly more slowly inside the membrane than its isotopically lighter counterpart. As a result, more of the heavier isotope is retained on the high-pressure side of the membrane during each run. The magnitude of isotopic fractionation generally increased with increasing water permeation rates. Additionally, experimental evidence shows that the FT-30 membrane exhibits essentially identical solute rejection with NaCl and LiCl under the same operating conditions.

Control of 36Cl production in carbonaceous shales by phosphate minerals

Abstract When using 36Cl to date very old groundwater in regional aquifer systems, knowledge of the subsurface 36Cl input into the aquifer system is essential. Although 36Cl can be produced through nuclear reactions in the subsurface, in many situations, the input of 36Cl into sedimentary aquifer systems by this avenue of production can be neglected. This is a valid assumption when investigating long-flowpath groundwater systems composed of sandstones, limestones, and shales of typical composition. These rock types are not sufficiently enriched in radioactive elements to produce significant 36Cl in the deep subsurface. Carbonaceous shales, on the other hand, can concentrate the radioactive elements necessary to produce significant 36Cl in the deep subsurface. Chlorine-36 ratios (36Cl/Cl) for a suite of Late Devonian and Pennsylvanian carbonaceous shales were calculated from bulk-rock chemistry as well as measured using accelerator mass spectrometry. The poor agreement between calculated and measured ratios is the result of the assumption of chemical homogeneity used by the calculation algorithm, an assumption that was not satisfied by the carbonaceous shales. In these shales, organic matter, clay minerals, and accessory minerals are heterogeneously distributed and are physically distinct on a micron-order scale. Although organic matter and clay minerals constitute the overwhelming bulk of the shales, it is the phosphate minerals that are most important in enhancing, and suppressing, 36Cl production. Minerals such as apatite and carbonate-apatite (francolite)—by including uranium, rare earth elements (REEs), and halogens—have an important impact on both neutron production and thermal neutron absorption. By incorporating both uranium and fluorine, phosphate minerals act as neutron production centers in the shale, increasing the probability of 36Cl production. By incorporating REEs and chlorine, phosphate minerals also act to shield 35Cl from the thermal neutron flux, effectively suppressing the production of 36Cl. To reconcile the measured 36Cl ratios with the ratios calculated assuming chemical homogeneity, the shales were artificially split into three fractions: organic, clay mineral, and phosphate mineral. Neutron production was calculated separately for each fraction, and the calculation results demonstrated that the phosphate fraction exerted much more control on the 36Cl ratio than the organic or clay mineral fractions. By varying the uranium and chlorine contents in the phosphate fraction, a new, heterogeneous 36Cl ratio was calculated that agreed with the measured ratio for the overwhelming majority of the carbonaceous shales. When using rock chemistry to calculate the 36Cl ratio, rock types that show mineralogical heterogeneity on a micron scale can be divided into bulk fractions and accessory fractions for separate calculations of neutron production and neutron absorption. In this manner, a more accurate, heterogeneous 36Cl ratio can be calculated for the rock as a whole.

Environmental Impacts of Acid Leachate Derived from Coal-Storage Piles upon Groundwater

Leachate emanating from a coal-storage area at an electricutility plant in Northwest Indiana (U.S.A.) is impacting groundwater quality. This assessment is based on results of along-term groundwater monitoring program conducted at Purdue Universitys Wade Utility Plant where a monthly average of 32,000metric tons of both high- and low-sulfur coal are stored. Groundwater from both a perched and major aquifer (the WabashValley Aquifer) as well as surface water from a large retentionpond were sampled monthly for 34 consecutive months. Such a long-term, continuous sampling scheme proved beneficial in identifying seasonal trends and anomalies within the chemistryof ground- and surface waters. Data show elevated concentrationsof sulfate and hardness in the perched zones and the WabashValley Aquifer (WVA). Lead in the WVA has also been reported ata concentration greater than the state maximum contaminant level, while several metals (Be, Cd, Cu, Pb, Ni, Se, and Zn) containedwithin retention-pond sediments have the highest concentrationsin aquatic sediments ever reported within the State of Indiana.Due to the buffering capacity of carbonate minerals in underlyingunconsolidated deposits, the acidic pH of coal-pile leachate is raised to values typical for groundwater in carbonate terrain common in Northwest Indiana. Further ameliorating the input of acid percolation is the dilution capabilities of both the WVA and the recharge pond. Hence, sites without such advantages would exhibit a greater degree of groundwater contamination than what is observed at Purdues coal-storage facility.