Applied Geochemistry (v.19, #11)

Water samples from short-screen monitoring wells installed along a 90-km transect in southwestern Kansas were analyzed for major ions, trace elements, isotopes (H, B, C, N, O, S, Sr), and dissolved gases (He, Ne, N2, Ar, O2, CH4) to evaluate the geochemistry, radiocarbon ages, and paleorecharge conditions in the unconfined central High Plains aquifer. The primary reactions controlling water chemistry were dedolomitization, cation exchange, feldspar weathering, and O2 reduction and denitrification. Radiocarbon ages adjusted for C mass transfers ranged from <2.6 ka (14C) B.P. near the water table to 12.8 ± 0.9 ka (14C) B.P. at the base of the aquifer, indicating the unconfined central High Plains aquifer contained a stratified sequence of ground water spanning Holocene time. A cross-sectional model of steady-state ground-water flow, calibrated using radiocarbon ages, is consistent with recharge rates ranging from 0.8 mm/a in areas overlain by loess to 8 mm/a in areas overlain by dune sand. Paleorecharge temperatures ranged from an average of 15.2 ± 0.7 °C for the most recently recharged waters to 11.6 ± 0.4 °C for the oldest waters. The temperature difference between Early and Late Holocene recharge was estimated to be 2.4 ± 0.7 °C, after taking into account variable recharge elevations. Nitrogen isotope data indicate NO3 in paleorecharge (average concentration=193 μM) was derived from a relatively uniform source such as soil N, whereas NO3 in recent recharge (average concentration=885 μM) contained N from varying proportions of fertilizer, manure, and soil N. Deep water samples contained components of N2 derived from atmospheric, denitrification, and deep natural gas sources. Denitrification rates in the aquifer were slow (5 ± 2× 10−3 μmol N L−1  a−1), indicating this process would require >10 ka to reduce the average NO3 concentration in recent recharge to the Holocene background concentration.

Since the 1980s, several in-lake restoration methods for eutrophied lakes have been developed based on artificially induced calcite precipitation and lime additions. The aim of these measures is to improve water quality and reduce the incidence of cyanobacterial blooms by increasing P adsorption and removal through natural calcite precipitation. The authors compared the relative efficiencies of lake marl application and natural calcite precipitation in immobilizing P, Mn and Fe in Lake Arendsee, northeastern Germany, and their subsequent value as a tool for managing eutrophication.Microprobe investigations of element distributions in calcite minerals of laminae precipitated annually during several years before and two years following the restoration suggest that recently grown biogenic calcites were more efficient at the co-precipitation and immobilization of P, Fe and Mn. The lake marl applied as 16% calcite/water mixture proved unsuitable as a restoration tool because of its coarse grain size mainly in the range of 20–1000 μm and associated organic matter which reduced its suspension time and sorption reaction rates in the water column. In Lake Arendsee or in similar eutrophic hardwater lakes with fast internal P recycling by rapid degradation of autochthonous organic matter in the water column and at the sediment surface (sedimentary TOC:TP ratios of 200 or more), biogenic calcites provide almost permanent P, Fe and Mn retention in the sediments. The study shows that epilimnetic additions of pure CaCO3 seed crystals comparable to natural calcite precipitation are the most effective for improving water quality. Small CaCO3 crystals exhibit a larger reactive surface, and thus, incorporate adsorbed dissolved nutrients and trace metals more rapidly during growth. However, in-lake water quality management strategies work best only if external point and non-point nutrient sources are reduced or diverted.

One of the French options for the final disposal of high-level radioactive waste is a deep geological repository. The Callovo-Oxfordian formation in the Meuse and Haute-Marne (France) was proposed, by the French agency for the management of radioactive wastes (ANDRA), to be the site of a research laboratory in order to evaluate in situ the interaction between the host formation and the engineered barriers.In the previous experimental part, secondary minerals such as zeolites, tobermorite and katoite have been observed after the alteration of the Callovo-Oxfordian clay under alkaline conditions. Three different reactions of phyllosilicates alteration have been proposed as a function of both pH and chemical composition of the reacting solutions. The aim of this work was to simulate the reaction between the Callovo-Oxfordian clay and high-pH solutions in order to validate the proposed alteration reactions from experimental data.The thermodynamic modelling of stability relationships among minerals in the Callovo-Oxfordian clay was performed using the solution speciation solubility modelling code KINDIS at 120 °C for different chemical media (NaOH, KOH and Ca(OH)2). Input data were the mineralogical composition of the <2 μm fraction of Callovo-Oxfordian clay at nearly the same burial depth of the underground laboratory construction as well as the neoformed minerals observed after the experimental alteration of this clay (zeolites and cement phases). In the NaOH and KOH runs, anacime, phillipsite and chabazite, respectively were the first stable minerals which appeared at the beginning of modelling. In the Ca(OH)2 runs, oversaturated with respect to portlandite, katoite and then tobermorite were the first minerals occurring at the beginning of the alteration. Smectites and illites were undersaturated whatever the chemical composition. Both minerals and the interstratified illite-smectite were less stable in Ca(OH)2 than in NaOH and KOH solutions. Concerning the secondary minerals, a discussion is developed in order to improve modelling and to predict the long term reactions between clays and alkaline solutions.

The rare earth elements (REEs) have been used as analogs for understanding the behavior of actinide elements in geological conditions. Here the authors discuss the usefulness of the change of Eu abundance in a fractured aquifer for deducing the behavior of Am under geological conditions. The REE abundance of fractured core samples from Precambrian crystalline rocks in the Namyangju area, Korea provides a good example of a water–rock interaction in the low temperature aqueous environment. The rock types studied are mainly biotite banded-gneisses with mineral assemblages dominated by biotite, K-feldspar, quartz and plagioclase. Calcite, chlorite, muscovite and sericite occur as secondary minerals, with calcite being the main filling material in fractures. In general, the core samples from 5 boreholes are enriched in light REE (LREE) and depleted in heavy REE (HREE), with (La/Yb)N=7.3 to 67.6 and have negative Eu anomalies. However, positive Eu anomalies also occur in the fractured core samples that contain secondary calcite. Results of leaching experiments show that a variation of chondrite-normalized REE patterns is associated with the fracture-filling calcite, where Eu has been reduced and selectively concentrated in the solutions from which calcite has precipitated. The cohesive energy of Am in actinide series is similar to that of Eu. Hence, the behavior of Eu in the fractured aquifer may be an important key to understanding and predicting the behavior of Am in the geological environment after radioactive waste disposal in the ground. In addition, a slight Yb anomaly from chondrite-normalized REE pattern can be observed by a calcite leaching test, which also suggests that extremely reducing conditions possibly existed in the past groundwater environment.

Magnesium and Ca concentration ratios, fulvic acid content, total dissolved inorganic carbon (DIC) and pH were determined in seepage water and drip water samples collected during one seasonal cycle between June 2000 and May 2001 above and in the Béke Cave of Aggtelek (Hungary). Seepage water samples were collected at 0.5 and 7 m below ground level from an observation point situated above the cave. Drip water was collected 40 m underground from a group of stalactites. The fulvic acid concentrations were determined by fluorescence spectrometry after pre-concentration on a XAD-8 chromatographic column. Calcium and Mg concentrations were measured by inductively coupled plasma atomic-emission spectrometry. DIC was determined with a CO2 – selective electrode. DIC values increased and the fulvic acid concentrations and Mg and Ca concentration ratios, generally, decreased with depth. The highest flux of fulvic acid was observed in spring. The fulvic acid flux increased by a factor of 2.6–3.6 and 1.4 for groundwater and drip water, respectively, compared with those registered in the winter samples. The variations in the Ca, Mg and fulvic acid concentrations of the seepage and drip water samples relate to the variable drip rate. The results revealed that there is a strong correlation between the daily average surface temperature, daily amount of precipitation and drip water rate registered in the cave.

Mercury and other geochemical parameters were measured in mine-waste calcines, stream sediments and surface waters from the Wanshan Hg district in the Guizhou Province of China to evaluate adverse effects of Hg to the surrounding environment. Wanshan ore is dominantly cinnabar, and lacks other sulfide minerals and as a result, heavy metals other than Hg are rare in the calcines and mine-water runoff. Host rocks in the Wanshan area are dominantly carbonates, which upon weathering, produces alkaline-water runoff. Surface waters emanating from calcine piles have water pH varying from 10.6 to 11.8, contain high EC values, and variable major element geochemistries. Dissolved Hg concentration decreases from 300 to 1900 ng/l in mine water to 72 ng/l in stream water collected significantly downstream from the mine. High concentrations of Hg in mine-waste calcines and alkaline surface water are the major threats to the surrounding environment. Alkalinity is also naturally attenuated downstream from the Wanshan mine. However, erosion and transport of Hg in sediment and water and highly alkaline mine-water runoff, may adversely affect sediment, aquatic and biological columns downstream from the Wanshan mine area.

Controls on water chemistry of the Pilcomayo river (Bolivia, South-America) by A.J.P. Smolders; K.A. Hudson-Edwards; G. Van der Velde; J.G.M. Roelofs (1745-1758).
In order to reveal the intra-annual variability of the major ion composition of the Pilcomayo river, a dryland river, and its relationship to discharge, water samples were taken at regular time intervals from May 1998 until February 1999 at the town of Villa Montes (Bolivia). Water chemistry of the Pilcomayo river was highly variable during the year and strongly influenced by differences in discharge between the wet and the dry season. Halite dissolution appeared to play an important role and both Cl and Na concentrations became very high (±10 mmol L−1) during the dry season. Pyrite weathering and dissolution of gypsum, dolomite and calcite determined Ca, Mg, CO3 and SO4 chemistry. At the onset of the rainy season `rinse out' effects occurred, resulting in marked concentration peaks especially for the least soluble ions. Possible effects on biota, such as consequences for trace metal toxicity, are discussed briefly.

Sediment capping in eutrophic lakes – efficiency of undisturbed calcite barriers to immobilize phosphorus by Ute Berg; Thomas Neumann; Dietfried Donnert; Rolf Nüesch; Doris Stüben (1759-1771).
The application of a calcite barrier was investigated to prevent P release from eutrophic lake sediments. For this, different calcite materials varying in grain size, specific surface area (SSA) and roughness of the surface were applied. Rohrbach calcite (RB) is a crushed Jurassic limestone with an SSA of 4.3 m2  g−1 and a microporosity of 7%. Merck calcite (M2066) is an analytical reagent with an SSA of 1 m2  g−1 and a microporosity of 17%. The industrially manufactured calcites U1 and U3 have higher SSAs of 20 and 70 m2  g−1 and a microporosity of 20% and 9%, respectively.Short-term batch experiments and long-term sediment incubation experiments showed that the calcite saturation of the water (SICaCO3 ) predetermines the P fixation mechanisms. In waters supersaturated with respect to calcite (SICaCO3  > 0), adsorption and simultaneous co-precipitation with calcite are the predominant processes for P fixation. In addition, precipitation of Ca–P compounds occurs on the surface of calcite seed crystals due to the decrease in the interfacial nucleation energy. The capacity of P fixation was greatly influenced by the physical properties of calcite grains. An increase in the SSA from 1 (M2066) and 4 m2  g−1 (RB) to 67 m2  g−1 (U3) improves the efficiency of P removal from about <5% (RB, M2066) to 90% (U3). The grain size affects the P fixation especially in waters close to equilibrium and of weak supersaturation, where small calcite grains dissolve and initiate supersaturation with respect to Ca–P compounds. Under these conditions SRP concentrations decrease by 30% (M2066), 60% (RB) and about 95% (U3) relative to their initial concentrations. At SICaCO3  < 0, calcite dissolution causes an increase in the dissolved Ca concentration and initiates Ca–P precipitation lowering the SRP concentrations by about 65% (M2066, RB) and about 100% (U3). This process is significantly enhanced by the surface roughness of the calcite grains (M2066) due to numerous micropores of <2 nm in diameter.The results showed that calcite barriers could be optimized individually in accordance to the hydrochemical conditions in lakes to increase the efficiency of P retention in sediments. An application of a 1 cm thick RB barrier resulted in an 80% reduction of the P flux from the sediment for at least 2–3 months, whereas a quartz sand barrier of 2–4.5 cm thickness containing 2 wt% of highly active calcite such as U3 and U1 quantitatively prevents release of P from eutrophic lake sediments for at least 7 and 10 months, respectively.

Constructed wetland systems often represent a low cost means for the remediation of dissolved metal and SO4-rich acid mine drainage. The surface water, sediment, and porewater geochemistry of one such wetland were investigated to determine the active geochemical processes that govern wetland performance. The influent drainage has a pH of 2.6, 200–250 mg/L of dissolved Fe, and Al, Mg, and Mn concentrations of 30–100 mg/L. Sulphate concentrations are also high, ranging from 500–1000 mg/L. Sediment cores show high levels of amorphous Fe oxyhydroxides near the surface and some amorphous reduced inorganic sulphur (RIS) in deeper layers. Taken together with a decline in porewater SO4 with depth, the data suggest that bacterial SO4 reduction is occurring. All sediment cores display an increase in porewater SO4 with depth in their top sections, often forming a subsurface peak, which could be attributed to oxidative dissolution of RIS at the redox boundary. However, the increase in SO4 is also accompanied by a rise in porewater pH. Desorption experiments yielded up to 8 wt% SO4 in these top layers suggesting that pH dependant sorption is an active SO4 removal mechanism along with the precipitation of Fe oxyhydroxysulphates, which also removes dissolved Fe. The maximum treatment efficiency of the wetland is 25% for the removal of dissolved Fe, and the retention of other species is equally poor. Comparison with other wetland systems suggests that poor performance is due to the low pH and high dissolved metal concentration of the drainage together with deterioration in the wetland over time, as a result of reduced active surface area and retention time.

Arsenic behaviour in gold-ore mill tailings, Massif Central, France: hydrogeochemical study and investigation of in situ redox signatures by F. Bodénan; P. Baranger; P. Piantone; A. Lassin; M. Azaroual; E. Gaucher; G. Braibant (1785-1800).
Historical Au-ore exploitation at the Chéni mine in the Massif Central, France, generated 525,000 tonnes of finely ground mill tailings deposited in a heap that has spread with time into three settling basins. The tailings, which are rich in quartz (80%), mica and clay minerals (10% of illite, smectite, kaolinite and chlorite), feldspars (5%) but poor in carbonates (<1%), also contain sulphides (around 5%, mainly pyrite and arsenopyrite). Arsenic content of the tailings is around 6 g kg. This paper describes the geochemistry of drainage waters, with special attention paid to in situ values of the three major redox couples, namely Fe(II)/Fe(III), As(III)/As(V) and S(IV)/S(VI). The water samples range from acidic and oxidized (pH 2.9, E h +700 mV) to moderate pH and weakly reducing (pH 7.6, E h 15 mV). The waters are rich in SO4 and Ca and have variable As (0.05–95 mg L−1) and Fe concentrations (0.07–141 mg L−1). Reduced As(III) species predominate over As(V) species (As(III)/As(V) up to 21), whereas oxidized forms of Fe and S are favoured (Fe(II)/Fe(III) up to 0.5, and S(IV)/S(VI) up to 1).Thermodynamic calculations were performed with the PHREEQC and EQ3NR codes based on a revised As database to evaluate saturation indices (SI) of the waters in relation to the main minerals and define which redox couples control the redox state of the system. The important role of carbonates, though only present in small amounts, explains the acid buffering generated by the oxidation of sulphides for waters in the pH 7–7.5 range. Measured E h appears to fall between the calculated E h of the Fe(II)/Fe(III) couple and that of the As(III)/As(V) couple, illustrating redox disequilibrium.

Metal accumulation in Baltimore Harbor: current and past inputs by Robert P. Mason; Eun-Hee Kim; Jeffery Cornwell (1801-1825).
The analysis of sediment cores taken from 3 regions of Baltimore Harbor, Maryland, USA, provide insight into contamination in the 20th century, as well as in earlier times. Overall, while there appears to have been a decrease in concentration for most metals over the last 20 a, concentrations of some metals in surface sediment layers, such as Cr and Pb, are still elevated. The deeper sediments, which reflect times before 1900, often contained higher concentrations of metals and reflect a large historical contamination of this region due to industrial inputs and other activities. Mining and smelting impacts are apparent, being reflected by elevated concentrations of Zn, Pb, Cu, Cd and As, and such impacts are seen in the Middle Branch of the estuary during the 1700s and elsewhere in the 1800s as Cu and other ore processing, and steel manufacture, increased. Comparison of the profiles for the metals, and accumulation rates, suggest that the sources of most metals are different. For each site, the particular metal signature recorded in the sediment can be explained in terms of the known activities in the watershed. At one site, sediments from the period prior to the European settlement were analyzed and these show low concentrations reflective of the situation before industrialization and urbanization of the watershed.

The springs of Lake Pátzcuaro: chemistry, salt-balance, and implications for the water balance of the lake by James L. Bischoff; Isabel Israde-Alcántara; Victor H. Garduño-Monroy; Wayne C. Shanks III (1827-1835).
Lake Pátzcuaro, the center of the ancient Tarascan civilization located in the Mexican altiplano west of the city of Morelia, has neither river input nor outflow. The relatively constant lake-salinity over the past centuries indicates the lake is in chemical steady state. Springs of the south shore constitute the primary visible input to the lake, so influx and discharge must be via sub-lacustrine ground water. The authors report on the chemistry and stable isotope composition of the springs, deeming them representative of ground-water input. The springs are dominated by Ca, Mg and Na, whereas the lake is dominated by Na. Combining these results with previously published precipitation/rainfall measurements on the lake, the authors calculate the chemical evolution from spring water to lake water, and also calculate a salt balance of the ground-water–lake system. Comparing Cl and δ 18O compositions in the springs and lake water indicates that 75–80% of the spring water is lost evaporatively during evolution toward lake composition. During evaporation Ca and Mg are lost from the water by carbonate precipitation. Each liter of spring water discharging into the lake precipitates about 18.7 mg of CaCO3. Salt balance calculations indicate that ground water input to the lake is 85.9 × 106 m3/a and ground water discharge from the lake is 23.0 × 106 m3/a. Thus, the discharge is about 27% of the input, with the rest balanced by evaporation. A calculation of time to reach steady-state ab initio indicates that the Cl concentration of the present day lake would be reached in about 150 a.

Iron monosulfide formation and oxidation processes were studied in the extensively drained acid sulfate soil environment of the Tweed River floodplain in eastern Australia. Porewater profiles of pH, Eh, SO4 2−, Fe2+, Fe3+, Cl, HCO3 , and metals (Cd, Co, Cr, Cu, Ni, Pb and Zn) were obtained using in situ dialysis membrane samplers (`peepers'). Concentrations of acid volatile S (AVS), pyrite, total S, reactive Fe, total and organic C, simultaneously extracted metals (SEMs) and total elemental composition by X-ray fluorescence, were determined on sediment samples. The oxidation of pyrite in the surrounding landscape provides a source of acidity, Fe, Al, SO4 and metals, which are exported into the drainage system where they accumulate in the sediments and porewaters. Negative porewater concentration gradients of SO4 2− and Fe2+, and large AVS concentrations in the sediments, indicate Fe monosulfides form rapidly under reducing conditions and consume acidity and metals. Oxidation of the sediments during previous drought episodes has resulted in the conversion of monosulfides and pyrite to oxidised Fe minerals and the release of acidity, SO4 2−, Fe3+, and metals to the surface waters. These formation and oxidation cycles show that Fe monosulfides play an important role in controlling water quality in the drainage system.

Environmental assessment of copper–gold–mercury mining in the Andacollo and Punitaqui districts, northern Chile by Pablo Higueras; Roberto Oyarzun; Jorge Oyarzún; Hugo Maturana; Javier Lillo; Diego Morata (1855-1864).
The Coquimbo region has been one of the richest producers of Cu, Au and Hg in Chile, and some of the deposits have been mined almost continuously since the 16th century. To assess the potential environmental contamination in this region, the authors measured the concentration of Cu, As, Cd, Zn and Hg in samples of stream and mine waters, stream sediments, soils, flotation tailings, and mine wastes in the Andacollo (Cu, Au, Hg) and Punitaqui (Cu–Au, Hg) districts. The concentration of Hg in the atmosphere in these districts were also measured. Although contamination is strongly controlled by the ore in each district, metal dispersion is modified by the degree of metallurgical processing efficiency as shown by the outdated Cu flotation system at Andacollo (stream sediments Cu 75–2200 μg/g). Conversely, more efficient procedures at Punitaqui resulted in less stream contamination, where stream sediments contained Cu ranging from 110–260 μg/g. However, efficient concentration by flotation of a given metal (e.g. Cu) may lead to the loss of another (e.g. Hg up to 190 μg/g in the tailings at Punitaqui), and therefore, to contamination via erosion of the tailings (downstream sediments Hg concentrations up to 5.3 μg/g). Continued use of Hg for Au amalgamation at Andacollo has led to significant contamination in stream sediments (0.2–3.8 μg/g Hg) and soils (2.4–47 μg/g Hg). Communities in this region are underdeveloped, and decades of inefficient treatment of flotation tailings and waste-rock stock piles has resulted in significant contamination of the surrounding landscape.