Applied Geochemistry (v.19, #5)

Sediment and water samples from 12 saline pans on the semi-arid west coast of South Africa were analysed to determine the origin of salts and geochemical evolution of water in the pans. Pans in the area can be subdivided into large, gypsiferous coastal pans with 79–150 g/kg total dissolved salt (TDS), small inland brackish to saline (2–64 g/kg TDS) pans and small inland brine (168-531 g/kg TDS) pans that have a layer of black sulphidic mud below a halite crust. The salinity of coastal pan waters varies with the seasonal influx of dilute runoff and dissolution of relict Pleistocene marine evaporite deposits. In contrast, inland pans are local topographic depressions, bordered on the north by downslope lunette dunes, where solutes are concentrated by evaporation of runoff, throughflow and groundwater seepage. The composition of runoff and seepage inflow waters is determined by modification of coastal rainfall by weathering, calcite precipitation and ion exchange reactions in the predominantly granitic catchment soils. Evaporation of pan waters leads to precipitation of calcite, Mg–calcite, dolomite, gypsum and halite in a distinct stratigraphic succession in pan sediments. Bicarbonate limits carbonate precipitation, Ca limits gypsum precipitation and Na limits halite precipitation. Dolomitisation of calcite is enhanced by the high Mg/Ca ratio of brine pan waters. Brine pan waters evolve seasonally from Na–Cl dominated brines in the wet winter months to Mg–Cl dominated brines in the dry summer months, when 5–20 cm thick halite crusts cover pan surfaces. Pan formation was probably initiated during a drier climate period in the early Holocene. More recent replacement of natural vegetation by cultivated land may have accelerated salt accumulation in the pans.

Here, a new technique for the determination of dissolved He isotope ratios in ground-waters is presented. This method is based on the extraction and subsequent equilibrium of dissolved gases in an added “host” gas phase. Ultra pure N2 is placed in glass flasks (250 cc), containing water samples, that were hermetically sealed after their collection. After shaking in an ultrasonic bath for 10 min, an aliquot of the separated gas phase was removed from the flask for MS analysis. 3He/4He ratios are measured by using a modified double collector mass spectrometer (VG 5400-TFT). Helium and Ne concentrations are calculated by comparing the partial pressures of masses 4 and 20 of the samples with those of the air-standard measured by a quadrupole mass spectrometer (QMS;VG Quartz). Using He and Ne equilibrium partitioning coefficients, it is possible to calculate the amount of gas originally dissolved in the water. The technique was tested on both air-saturated waters (ASW) and thermal waters from Stromboli (Aeolian Islands, South Italy), the results of which confirmed good reproducibility (≌5%) and accuracy (≌3%) of the data. The method was then applied to three thermal water samples collected from the same volcanic area and the results compared with those of a fumarolic and a soil gas. The isotope ratios for dissolved He gave values of 4.06–4.23 Ra, which are significantly higher than those previously reported in the literature (3.0, 3.5 and 2.9 Ra) and that measured at the fumarole (3.09 Ra), suggesting a newer and higher isotopic signature for the volcanic system. The proposed method appears to be a useful tool in the determination of 3He/4He ratios in ground-water systems, especially when free gases are not available or are dangerous to collect.

A study of the geochemistry of fracture filling calcite from the Palmottu research site, a U–Th deposit located about 100 km NW of Helsinki, Finland, was part of a comprehensive natural analogue research project considering the migration of U in the crystalline bedrock environment. Fracture filling minerals act as records of the paleohydrological and chemical conditions and therefore provide insight into magmatic processes, ore deposits and long-term stability of the crystalline bedrock environment in relation to deep disposal of radioactive waste. An earlier investigation of calcite and associated U minerals identified a recent origin for these fracture infillings. Uranium series disequilibrium data indicated that relatively recent, glacial water had penetrated to a depth of 62 m. A fluid inclusion study, combined with the isotope geochemistry of several generations of fracture calcite, was undertaken to further study this aspect and to better understand the thermal and fluid history in the crystalline rock environment in general. The study revealed that at least 3 fluids were recorded by fracture calcites: 1) A crystalline calcite precipitated at 139–238 °C from a low salinity, Na–Cl fluid of magmatic or metamorphic origin, 2) A massive calcite (with high salinity fluid inclusions) precipitated at 136–141 °C from a high salinity, Ca–Na–Cl fluid of magmatic or metamorphic origin and 3) A massive calcite (with low salinity fluid inclusions) precipitated at 43–286 °C from a low salinity, Na–Cl fluid of magmatic or metamorphic origin that underwent equilibrium fractionation during cooling in a rock dominated system. The preservation of high temperature infillings and the cooling trend, indicating isotopic re-equilibration of water due to a very low water/rock ratio, demonstrate that the downward percolation of geologically recent waters is limited, despite the reworking of highly soluble U compounds in the upper tens of metres of the bedrock that was shown by U series disequilibrium studies. The veins in which these calcites occur have not experienced a later phase of fluid activity that dissolved or altered the calcites. Therefore, it is likely that they have not been reactivated as transmissive fractures since that initial hydrothermal episode, even during the Quaternary history in which the formation was subjected to ice sheet loading and unloading.

The acid neutralisation capacity (ANC) of a rock sample containing significant amounts of Fe carbonates, as conducted to determine net acid production potential (NAPP), can be a difficult parameter to determine. Various ANC tests are available to determine the ANC of carbonates. This work does not attempt to create another ANC test protocol; rather, it provides a refinement for existing tests. Results showed that a significant lag period may be needed (up to 432 h) after standard Sobek-type ANC tests for the complete hydrolysis of Fe associated with the ANC testing of siderite. This lag occurred even with standard industry modifications that include the addition of 2 drops of H2O2 at pH 4.5 during the back-titration. In this work the authors used a modification to the Sobek ANC test (the Modified Sobek ANC test) that included filtering and the addition of H2O2 at pH 4.5. This test was further modified by the continuance of this H2O2 addition (the H2O2 ANC test) until there was no subsequent pH drop (which is due to Fe hydrolysis reactions), thereupon the back titration was continued to pH 7.0. Results indicated that the ANC for siderite (after 0 h) using the H2O2 ANC test was similar to the ANC determined after 432 h lag by the Modified Sobek ANC test. This modification reduces the uncertainty related to static-test results for samples containing Fe carbonates. The test is simple to use, has industry application, and provides a better indication of the NAPP. The Modified Sobek ANC results for calcite and dolomite and the H2O2 ANC test for siderite were in good agreement with the mineralogical carbonate ANC (ANCcarb). ANCcarb was determined by calculation based on electron probe micro-analysis. Although lower than both the ANCcarb and the ANC determined by titration, the chemical ANC calculated from the ions present in the ANC digestion liquor also provided a good indication of the overall acid neutralisation capacity of the sample.

Geochemistry of mud volcano fluids in the Taiwan accretionary prism by Chen-Feng You; Joris.M. Gieskes; Typhoon Lee; Tzen-Fu Yui; Hsin-Wen Chen (695-707).
Taiwan is located at the collision boundary between the Philippine Sea Plate and the Asian Continental Plate and is one of the most active orogenic belts in the world. Fluids sampled from 9 sub-aerial mud volcanoes distributed along two major geological structures in southwestern Taiwan, the Chishan fault and the Gutingkeng anticline, were analyzed to evaluate possible sources of water and the degree of fluid-sediment interaction at depth in an accretionary prism. Overall, the Taiwanese mud volcano fluids are characterized by high Cl contents, up to 347 mM, suggesting a marine origin from actively de-watering sedimentary pore waters along major structures on land. The fluids obtained from the Gutingkeng anticline, as well as from the Coastal Plain area, show high Cl, Na, K, Ca, Mg and NH4, but low SO4 and B concentrations. In contrast, the Chishan fault fluids are much less saline (1/4 seawater value), but show much heavier O isotope compositions (δ18O=5.1–6.5 ‰). A simplified scenario of mixing between sedimentary pore fluids and waters affected by clay dehydration released at depth can explain several crucial observations including heavy O isotopes, radiogenic Sr contents (87Sr/86Sr=0.71136–0.71283), and relatively low salinities in the Chishan fluids. Gases isolated from the mud volcanoes are predominantly CH4 and CO2, where the CH4–C isotopic compositions show a thermogenic component of δ13C=−38 ‰. These results demonstrate that active mud volcano de-watering in Taiwan is a direct product of intense sediment accretion and plate collision in the region.

A dual-isotope approach to the nitrogen hydrochemistry of an urban aquifer by Teppei Fukada; Kevin M. Hiscock; Paul F. Dennis (709-719).
The potential for exploitation of urban aquifers is partly dependent on understanding the distribution and fate of urban N sources, such as sewage and fertilisers, that can limit the use of groundwater for public supplies. To investigate the application of the dual-isotope approach to understanding the N hydrochemistry of urban groundwater, this paper presents δ 15N–NO3 and δ 18O–NO3 data collected from two multi-level piezometers in the Sherwood sandstone aquifer beneath Nottingham in the English Midlands, UK. At one multi-level piezometer (Old Basford), depth sample measurements of δ 15N–NO3 in the range +9.2 to +11.4 ‰ and δ 18O–NO3 in the range +8.2 to +10.9‰, together with NO3 nitrate concentrations from 31.7 to 66.7 mg/l, are evidence for nitrification of sewage-derived inputs. In contrast, at the other multi-level piezometer (the Meadows), isotopically enriched samples (δ 15N–NO3 in the range +24.3 to +42.2 ‰ and δ 18O–NO3 in the range +20.5 to +29.4‰) are evidence for denitrification, although the compositional range of δ 15N–NO3 does not identify the N source without corroborating data. For the Meadows location, a cross-plot of δ 15N–NO3 versus δ 18O–NO3 gave an enrichment of the 15N isotope relative to the 18O isotope by a factor of 1.9, within the range of 1.3–2.1 reported for denitrification in other studies. This study has shown that the dual-isotope approach provides improved understanding of N sources and fate in the urban environment but further work is required to identify nitrification pathways to provide more confidence in the application and interpretation of δ 18O–NO3 measurements.

Hybrid empirical—theoretical approach to modeling uranium adsorption by Larry C. Hull; Christopher Grossman; Robert A. Fjeld; John T. Coates; Alan W. Elzerman (721-736).
An estimated 330 metric tons of U are buried in the radioactive waste Subsurface Disposal Area (SDA) at the Idaho National Engineering and Environmental Laboratory (INEEL). An assessment of U transport parameters is being performed to decrease the uncertainty in risk and dose predictions derived from computer simulations of U fate and transport to the underlying Snake River Plain Aquifer. Uranium adsorption isotherms were measured for 14 sediment samples collected from sedimentary interbeds underlying the SDA. The adsorption data were fit with a Freundlich isotherm. The Freundlich n parameter is statistically identical for all 14 sediment samples and the Freundlich K f parameter is correlated to sediment surface area (r 2=0.80). These findings suggest an efficient approach to material characterization and implementation of a spatially variable reactive transport model that requires only the measurement of sediment surface area. To expand the potential applicability of the measured isotherms, a model is derived from the empirical observations by incorporating concepts from surface complexation theory to account for the effects of solution chemistry. The resulting model is then used to predict the range of adsorption conditions to be expected in the vadose zone at the SDA based on the range in measured pore water chemistry. Adsorption in the deep vadose zone is predicted to be stronger than in near-surface sediments because the total dissolved carbonate decreases with depth.

The authors report results from an experimental study on mixtures of pure endmembers of natural clay and carbonate. The scientific rationale is an evaluation as to what extent B contents and B isotopes of carbonate samples may be obscured as a result of contamination with clay, particularly since both authigenic carbonates and biogenic carbonates (e.g. microfossil tests) often contain some clay.Three aliquots of a series of samples (each containing 0, 20, 40, 60, 80, 100% clay) were analyzed. Set 1 was washed with distilled, de-ionized water; for set 2 the HCl soluble parts were dissolved in 2 M HCl after washing; set 3 was completely digested with 30M HF prior to a series of ion exchanges. Isotope data of the endmembers are 6.6‰ (100% marble) and −4.6‰ (100% clay), with the clay being the dominant B source (ca. 50 ppm compared with <2 ppm in the carbonate).For set 1, only the B adsorbed to the clay was washed out with the de-ionized water (δ11Badsorbed=12.9–14.1‰±0.5‰), while no B was mobilized from the carbonate. The HCl-dissolvable B in washed samples of set 2 show slightly increasing B contents with higher clay contents, suggesting that dissolution of the marble as well as B mobilization from the clay account for this trend. δ11B isotopes tend towards more negative values when clay content increases, indicating that some structurally-bound B is lost from the sheets of linked (Si, Al)O4 tetrahedra of the clay mineral. This result shows that not only B adsorption, but possibly diffusion or weathering of broken edges of clay minerals releases some structurally bound B of clay minerals. Set 3, where bulk samples were completely HF-digested, shows as expected a linear increase in B concentrations and decreasing δ11B ratios with increasing clay content. The overall results suggest that relatively small amounts of clay (e.g. as contamination in a microfossil test) have no significant impact on the B content and δ11B measured for the carbonate, but that care has to be taken if clay exceeds 10wt.% (e.g. carbonate concretions, chimneys, etc.).

Distribution of some trace metals in Syrian phosphogypsum by M.S. Al-Masri; Y. Amin; S. Ibrahim; F. Al-Bich (747-753).
Distribution of Cu, Cd, Zn and U in a Syrian phosphoric acid plant byproduct, phosphogypsum, has been determined. Uranium, Cd, Zn and Cu were found to be more enhanced in small phosphogypsum particles (45–75 μm) where the highest concentration was found for Cu (51.7 ppm). In addition, the element transfer factors ( Trace element concentration in phosphogypsum (mg/kg) Trace element concentration in phosphate rock (mg/kg) ×100) from Syrian phosphate rock to phosphogypsum were calculated and found to be 30, 8 and 17% for Zn, Cd and U, respectively. Moreover, laboratory leaching experiments of phosphogypsum by distilled water, dilute H2SO4 solutions and selective extractants have been performed. Leaching results have shown that around 20% of the U and 100% of the Zn are transferred to the aqueous phase. Batch-wise leaching with dilute H2SO4 solutions shows increased solubility of U, Zn, Cu and Cd from phosphogypsum, while leaching with selective extractants has been performed to determine the amount of exchangeable trace metals which are adsorbed, on gypsum particle surfaces, the amount of trace metals present inside the gypsum lattice, the amount of trace metals associated with organic materials and the amount of trace metals soluble in acids. The results obtained in this study can be utilized to verify the environmentally safe use of phosphogypsum as an amendment to agricultural soils.

The chemistry of surface and ground waters in the Ngorongoro Crater, Tanzania, home to thousands of large mammals and a World Heritage Site, is controlled by the volcanic host rock lithology, evaporative concentration, mineral precipitation and redissoluton, and biological factors. Three groups of waters are informally differentiated based on their ranges of concentration: (1) dilute inflow (meteoric runoff and springs from short flowpaths: pH<8, Cl<10 mg/l); (2) concentrated inflow (concentrated runoff and springs from long flowpaths: pH=8–9, Cl=10–100 mg/l); and (3) brackish waters (pools and Lake Makat: pH>9, Cl>100 mg/l). Evaporative concentration and biological activity in swamps commonly produce strong geochemical gradients between dilute sources and peripheral concentrated ephemeral wetlands. Dilute inflow is found in the Lerai, Munge, and Oljoro Nyuki streams, and several large springs near the Crater wall such as Ngoitokitok and Seneto. Concentrated inflow is found in downstream reaches of the Munge stream, discharging from springs away from the Crater wall such as Engitati and Mti Moja, and in dry-season pools. Brackish waters are found discharging from springs on the southern margin of Lake Makat, in mudflats surrounding marshes, in ephemeral pools, and in the lake itself. Although few hydrologic data are available, the persistence of relatively fresh water in vegetated wetlands is consistent with lower sedge-dominated wetland evapotranspiration rates compared with open water. This suggests that wetlands may play an important role in ensuring fresh water availability in the basin, and it demonstrates the need for future hydrologic study. Most of the Ngorongoro waters originate as rainfall outside the Crater, and travel into the basin as surface or ground water flow, emphasizing the need for a watershed-scale approach to land management.

Anthropogenic components of heavy metal (Cd, Zn, Cu, Pb) budgets in the Lot-Garonne fluvial system (France) by Stéphane Audry; Jörg Schäfer; Gérard Blanc; Cécile Bossy; Gilbert Lavaux (769-786).
Heavy metal (Zn, Cd, Cu and Pb) mass balances in the Lot-Garonne fluvial system have been established for 1999 and 2000. The mean annual discharges of these years are close to the mean discharge of the previous decade. The estimated annual dissolved and particulate fluxes in this model watershed integrate daily input from diffuse and point sources, diffusive fluxes at the water/sediment interface, changes in the dissolved-particulate partition and changes in sediment stock. Cadmium, Zn, Cu and Pb entering the Gironde estuary via the Garonne River (11–14 t a−1 of Cd; 1330–1450 t a−1 of Zn; 126–214 t a−1 of Cu and 127–155 t a−1 of Pb) are mainly transported in the particulate phase and the major part (i.e. ∼74 to 96% for Cd, ∼60% for Zn, ∼50 to 60% for Cu and ∼80% for Pb) is transported by the Lot River. The main anthropogenic heavy metal point source is located in a small upstream watershed (Riou-Mort River) accounting for at least 47% (Cd), ∼20% (Zn), ∼4% (Cu) and ∼7 to 9% (Pb) of the total heavy metal inputs into the Garonne River, although it contributes only 1% of the discharge. Mass balances for 1999 suggest that under mean annual hydrologic conditions on the basin scale, the heavy metal budget of the Lot-Garonne fluvial system is balanced and that the stocks of Cd [200 t; Environ. Tech. 16 (1995) 1145] and Zn in the Lot River sediment are constant under mean discharge conditions. Heavy metal input by molecular diffusion at the sediment surface represents an important component of dissolved metal inputs into the system (e.g. 30% for Cu). Except for Cu, these dissolved inputs are totally removed from solution by SPM. Based on the generally constant Zn/Cd (∼50) concentration ratio in sediment cores from the polluted Lot River reaches and the sediment stock of Cd [200 t; Environ. Tech. 16 (1995) 1145], the present day Zn stock in the Lot River sediments has been estimated at about 10,000 t. In addition to the mobilization of river-bed sediment and associated heavy metals by intense floods, local human activities, including river-bed dredging, may strongly modify the heavy metal budget of the river system. In 2000, the dredging-related remobilization of polluted Lot River sediment released 2–6 t Cd. This additional Cd point source was estimated to account for 15–43% of the gross inputs into the Gironde Estuary.

Geochemistry of high-pH waters from serpentinites of the Gruppo di Voltri (Genova, Italy) and reaction path modeling of CO2 sequestration in serpentinite aquifers by Francesco Cipolli; Barbara Gambardella; Luigi Marini; Giulio Ottonello; Marino Vetuschi Zuccolini (787-802).
The large number of geochemical data gathered on the Gruppo di Voltri springs confirm that progressive interaction of meteoric waters with ultramafic rocks variably affected by serpentinization leads initially to the formation of Mg–HCO3 waters when the system is open to CO2, and Na–HCO3 and Ca–OH type water upon further interaction with the rock, under highly reducing closed-system conditions with respect to CO2. As indicated by 3H data, these high-pH waters have had long residence times underground in deep aquifers hosted by serpentinitic rocks. These waters are the only available evidence of the presence of such deep aquifers. High-pressure injection of CO2 into these deep aquifers was simulated by reaction path modeling. Results indicate that this is a feasible methodology to reduce the inputs of anthropogenic CO2 into the atmosphere. Serpentinitic rocks have a high capacity for CO2 sequestration, mainly through formation of carbonate minerals. Dissolution of serpentinitic rocks and precipitation of magnesite and silica minerals occurs naturally in areas of high terrestrial CO2 fluxes such as in southern Tuscany, corroborating the feasibility of this methodology of CO2 sequestration. However, this process causes a progressive decrease in the porosity of the aquifer, at least under closed-system conditions. These side effects must be carefully evaluated by means of further laboratory tests and field activities.

Metal attenuation at the abandoned Spenceville copper mine by Mara Ranville; Daniella Rough; A. Russell Flegal (803-815).
Acid mine drainage from the abandoned Spenceville Cu mine contributes toxic levels of metals to nearby streams. The majority of metal contamination, however, is removed from the system over a short spatial distance. The primary mechanism responsible for this geochemical scavenging is adsorption onto Fe, Mn and Al-hydroxide precipitates that are clearly visible in the streambed. A 5-step sequential extraction procedure was performed on the streambed sediments in order to characterize solid phase speciation of metals in this area. Two fractions (carbonate and amorphous Fe-Mn hydroxide) were considered to be bioavailable, with the possibility of the metals in these phases being remobilized under changing ambient conditions. Copper and Zn were found to have the highest bioavailability, followed by Cd, Pb and Al. Although most metal in the acid mine drainage was rapidly removed from the system, aqueous concentrations of Cu and Zn exceeded California water quality criteria 0.5 km downstream of the mine site.