Applied Geochemistry (v.20, #7)

The use of wetlands to treat mine effluent has grown in popularity over the past two decades, although the processes by which the natural systems function are often poorly understood. This field-scale investigation utilises daily data over a 9-month period in assessing the processes leading to the remediation of mine effluent within a natural wetland on the Zambian Copperbelt. The study differentiates effluent remediation through dilution from pollutant retention. Decreased wetland outflow concentrations of SO4 and Na are due to dilution only, while Co (50%) and especially Cu (83%) are retained within the wetland. Retention was linked to adsorption onto new or primed surfaces during an initial period of effluent release into the system and to processes related to pH buffering to 7.5. The wetland’s acid buffering capacity was largely the result of carbonate-rich groundwater discharge into the wetland. Although this buffering capacity likely shows little seasonal fluctuation (20–80 kmol/day), the impact of acidic effluent input on the wetland itself probably varies markedly between seasons, owing to the temporal and spatial characteristics of discharge from the catchment’s aquifers. Assessment of other natural wetlands in the region indicated that some (circa 15%) showed similar catchment size, hydrochemical and hydrogeological characteristics as those of the New Dam wetland, likely demonstrating a similar effluent remediation potential as that described here.

Distributions of 21 major and trace elements in HNO3 extracts of different horizons were studied in 13 podzol profiles from the boreal forest in different parts of Norway using ICP–MS. On the basis of ratios between the HNO3-extractable fractions in the various horizons some general trends were elucidated. Two different groups of elements concentrated in the humus layer relative to the mineral horizons were identified, one mainly associated with contributions from air pollution (As, Cd, Sb, Pb), another one with plant nutrient circulation (K, Ca, Mn and to a lesser extent Mg, Co, Ni, Rb) and some with both mechanisms (Cu, Zn, Tl). The elements most clearly enriched along with Fe in the B horizon were V, Pb, Al, and Cr in that order, Pb partly because of leaching from the polluted organic surface soil. Four soils in the far south showed a behaviour distinctly different from the rest and were treated as a separate group. Relative to the more northerly sites the surface horizons of these soils were strongly depleted in lithogenic elements (Mg, Al, K, Ca, Sc, V, Cr, Mn, Fe, Co, La) and enriched in elements typical of long-range transport of pollutants (As, Cd, Sb, Tl, Pb). Also the B horizon in the southern soils was strongly depleted in the lithogenic group elements, including Fe and the associated metals. The main reason for this difference is assumed to be the greater influence of transboundary air pollution and associated metals and stronger soil acidification in the far south of the country.

A geochemical model is developed for the immobilization of U in the presence of metallic Fe. Zero-valent iron (ZVI) serves as a reducing agent inducing the reductive-precipitation of U, and ZVI corrosion products can serve as absorbing agents for U. The numerical model developed allows the complex interactions of U in solution in differing concentrations to be examined, under variable pH and redox conditions, with or without carbonate, in the presence of ZVI of different size and surface area. It incorporates Fe corrosion, Fe(II) and Fe(III) corrosion product formation, reductive-precipitation of U from the soluble U(VI) valence to the poorly soluble U(IV) valence, adsorption/de-sorption of U onto the Fe oxide corrosion products, and aqueous speciation. The processes of Fe corrosion and reductive precipitation of U are simulated as non-equilibrium, an improvement over other geochemical models. The reductive-precipitation process may use either ZVI or Fe(II) as the reducing agent. The model is calibrated using 3 separate sets of experimental data from published literature that cover a wide range of redox conditions. Sensitivity of the model predictions to variations in input parameters is examined. The simulation results show that the different published experimental results can be explained by different solution chemistries in the studies, specifically O2 and CO2 availability and pH, and the amount and surface area of the metallic Fe. With this numerical model the behavior of U in ZVI containing systems over a range of conditions realistic for groundwater can be investigated. By synthesizing observations across several experimental studies, it will lead to a broader understanding of the processes controlling U immobilization under varied geochemical conditions.

The dissolution of high-FeO olivine rock from the Lovasjärvi intrusion (SE-Finland) at 25 °C as a function of pH by Lara Duro; Fatima El Aamrani; Miquel Rovira; Javier Giménez; Ignasi Casas; Joan de Pablo; Jordi Bruno (1284-1291).
The high-FeO olivine-rich rock from the Lovasjärvi intrusion (65% olivine, 20% plagioclase, 8% magnetite, 4% pyroxene and 3% serpentine) has been proposed as a potential redox-active backfill-additive in deep high level nuclear waste repositories. In this work, the authors report on kinetic dissolution studies of this solid under different pH and redox conditions performed by using a flow-through methodology. Assuming that silicon is mainly released to solution from the olivine contained in the solid, the experimental results have been adjusted to an empirical rate law as a function of proton concentration. The proton concentration reaction orders agree with results found in the literature for both acidic and alkaline pH ranges. The calculations conducted with the reactive transport code RETRASO show that at alkaline pH, the olivine rock might have a lower redox buffer capacity than expected.

The Kouris catchment is located in the south of the Troodos massif in Cyprus. It constitutes one of the biggest catchments of the island with important freshwater resources. Geologically, the catchment includes an ophiolitic complex outcropping in the north which is overlaid by sedimentary rocks in the south. The hydrology is driven by a Mediterranean climate, a mountainous topography, and a complex distribution of the hydrogeological properties resulting from the complex geology.To improve the understanding of groundwater hydrology of the Kouris catchment, 176 groundwater and precipitation samples were collected and their 3H contents were analyzed. The three-dimensional 3H transport in the groundwater was simulated by the PMPATH code. For numerical modelling, a regional input function of 3H in precipitation was constructed from a linear regression between data for Cyprus and for neighboring meteorological stations. The calculated residence times for the groundwaters in the sedimentary aquifer and Pillow Lavas were greater than 48 a and were considerably greater than those of the ophiolitic complex (14–30 a). The calibrated aquifer porosities were in a range of 0.05–0.06. The PMPATH model was applied for delineation of spring catchments that were represented by quite narrow zones of lengths up to 5 km.Another contribution resulting from the 3H analysis was a better understanding of the river–aquifer interactions. In most of the southern part, the lithified sediments received only negligible amounts of water from the rivers, while the alluvial aquifer contained mostly water infiltrated from rivers. The largest springs in the southern part, associated with the alluvial aquifer, also discharged water identical to that in the rivers.

Impact of co-combustion of petroleum coke and coal on fly ash quality: Case study of a Western Kentucky power plant by James C. Hower; Gerald A. Thomas; Sarah M. Mardon; Alan S. Trimble (1309-1319).
Petroleum coke has been used as a supplement or replacement for coal in pulverized-fuel combustion. At a 444-MW western Kentucky power station, the combustion of nearly 60% petroleum coke with moderate- to high-sulfur Illinois Basin coal produces fly ash with nearly 50% uncombusted petroleum coke and large amounts of V and Ni when compared to fly ash from strictly pulverized coal burns. Partitioning of the V and Ni, known from other studies to be concentrated in petroleum coke, was noted. However, the distribution of V and Ni does not directly correspond to the amount of uncombusted petroleum coke in the fly ash. Vanadium and Ni are preferentially associated with the finer, higher surface area fly ash fractions captured at lower flue gas temperatures. The presence of uncombusted petroleum coke in the fly ash doubles the amount of ash to be disposed, makes the fly ash unmarketable because of the high C content, and would lead to higher than typical (compared to other fly ashes in the region) concentrations of V and Ni in the fly ash even if the petroleum coke C could be beneficiated from the fly ash. Further studies of co-combustion ashes are necessary in order to understand their behavior in disposal.

Acid mine drainage in the Iberian Pyrite Belt (Odiel river watershed, Huelva, SW Spain): Geochemistry, mineralogy and environmental implications by Javier Sánchez España; Enrique López Pamo; Esther Santofimia; Osvaldo Aduvire; Jesús Reyes; Daniel Barettino (1320-1356).
This work reports the physical properties and water chemistry of 64 AMD discharges from 25 different mines in the IPB draining to the Odiel river watershed, which have been investigated during the hydrologic year 2003–2004. These AMD solutions vary largely in flow rate and chemical composition both spatially (between the different mine sites, suggesting a strong geologic control on AMD chemistry) and seasonally (due to marked hydrologic variations), and include cases with very low pH (mostly in the range 1.4–4), and extreme sulphate (up to 44 g/L SO 4 2 - ) and metal content (e.g., up to 7.7 g/L Fe, 2.6 g/L Al, or 1.4 g/L Zn). Different hydrogeochemical facies of AMD (namely, Fe(II)/anoxic, Fe(III)/suboxic, and aluminous/oxic) are recognized in the field, as a response to the continuous oxidation and hydrolysis of dissolved Fe. Relevant geochemical aspects of these AMD environments are discussed, including: (i) the redox chemistry of the Fe(II)/Fe(III) couple, (ii) the reaction rates for bacterially catalyzed oxidation of Fe(II) and hydrolysis of Fe(III), (iii) the role played by dissolved Fe and Al in the acidity and chemical buffering of the AMD systems, and (iv) the solubility and trace metal retention capacity of the Fe oxyhydroxysulphate and hydrated sulphate minerals commonly associated with AMD. In addition, the mineralogy and chemistry of the Fe precipitates (schwertmannite, jarosite, goethite, ferrihydrite), Al phases (e.g., basaluminite) and Mg–Fe–Al efflorescent SO4 salts (e.g., epsomite, hexahydrite, copiapite, halotrichite, rozenite, coquimbite) present in the AMD-generating mine sites, have also been studied. The mineralogy of the Fe precipitates is well correlated with the water pH (with jarosite at pH ∼ 2, schwertmannite at pH 2–4, basaluminite at pH 4–5, and ferrihydrite at pH > 6). Schwertmannite appears to be the most important mineral phase, both in controlling the Fe solubility at pH 2–4, and as sorbent of trace elements (As, Cu, Zn), which favours natural attenuation. Finally, a basin-scale environmental perspective is given in order to evaluate the impact of AMD on the water quality, including calculation of metal loadings transported by AMD from the most important mine districts in the province.

Analysis of precipitates from reactions of hyperalkaline solutions with soluble silica by Kholoud Mashal; James B. Harsh; Markus Flury; Andrew R. Felmy (1357-1367).
Cancrinite, sodalite, and zeolite A have been found to form upon contacting hyperalkaline simulated tank waste (STW) with vadose zone sediments from the Hanford Reservation. Here, soluble silica and STW are used to study mineral formation and transformation. Two Hanford sediment fractions (diameters <50 and >50 μm instead of soluble silica) are also used as silica sources for comparison. A series of batch experiments at 50 °C and 25 days duration were conducted by reacting 0.026 mol/kg soluble Si with 6 different STW solutions. The STW solutions differed in NaOH and Al concentrations. Cancrinite, sodalite, and zeolite A formed when soluble Si was used as the Si source. The minerals were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FT-IR), and 27Al and 29Si magic-angle spinning nuclear magnetic resonance (MAS-NMR). Larger NaOH and Al concentrations favored formation of the more compact structures of cancrinite and sodalite. At larger NaOH concentration more Al for Si substitution occurred in the tetrahedral sites. A greater Al(4)/Al(6) ratio in the solids was found for the higher Si/Al ratio solutions based on NMR results. Mixtures of cancrinite and sodalite were characterized by particles with lepispheric morphology. At low Al concentration, increasing NaOH resulted in distinct hexagonal, prismatic particles common to crystalline cancrinite. At low Al/Si ratio, the characteristic cubic morphology of zeolite was observed in addition to cancrinite and sodalite.

Geochemical modeling was used to investigate the fate of major constituents and Cr and other trace elements when acid waters from the derelict Libiola mine are mixed with the Gromolo creek waters, and related processes. This was carried out in two steps: (1) Mixing between acid waters and stream waters was simulated using the EQ3/6 software package; the results showed that mixing brings about the precipitation of some secondary solid phases, mainly a ferrihydrite-rich solid mixture of hydroxides and basaluminite. (2) The sorption of cations (Cr3+, Cu2+, Zn2+, Mn2+, Ni2+, Ca2+, and Mg2+) and anions ( CrO 4 2 - and SO 4 2 - ) on ferrihydrite was modeled by means of the code MINTEQA2, treating the concentrations of relevant solutes and the moles of ferrihydrite (computed by means of EQ3/6) as input data.The results of this two-step geochemical modeling generally agree with analytical data, supporting the following conclusions: (a) dissolved Mg, Ca, Na, SO4, Cl, HCO3 and SiO2 are mainly controlled by acid water–stream water mixing; (b) Fe is mainly incorporated into ferrihydrite and subordinately in schwertmannite and jarosite; (c) Al is chiefly sequestered in basaluminite and alunite; (d) K is temporarily incorporated into jarosite and alunite and is released into the aqueous solution upon their dissolution; (e) the mobility of Cu, Zn, Ni, Mn, and Cr(VI) is mainly controlled by sorption on precipitating ferrihydrite; (f) Cr(III) is immobilised by both incorporation into ferrihydrite and sorption on its surface.The effects of acid mine waters on the Gromolo aqueous system are not ameliorated to an acceptable extent by natural attenuation through dilution and related processes.

Hydrological and biogeochemical dynamics of the minor and trace elements in the St. Lawrence River by Bernard Rondeau; Daniel Cossa; Pierre Gagnon; Thanh T. Pham; Charline Surette (1391-1408).
Surface water samples from the St. Lawrence River were collected in order to study the processes controlling minor and trace elements concentrations (Al, Fe, Mn, Cd, Co, Cu, Ni and Zn), and to construct mass balances allowing estimates of the relative importance of their natural and anthropogenic sources. The two major water inputs, the upper St. Lawrence River, which drains waters originating from the Lake Ontario, and the Ottawa River were collected fortnightly over 18 months. In addition, other tributaries were sampled during the spring floods. The output was monitored near Quebec City at the river mouth weekly between 1995 and 1999. Dissolved metal concentrations in the upper St. Lawrence River carbonated waters were lower than in the acidic waters of the tributaries draining the crystalline rocks of the Canadian shield and the forest cover. Biogeochemical and hydrodynamic processes occurring in Lake Ontario drive the seasonal variations observed in the upper St. Lawrence River. Biogeochemical processes relate to biological uptake, regeneration of organic matter (for Cd and Zn) and oxyhydroxide formation (for Mn and Fe), while hydrodynamic processes mainly concern the seasonal change in vertical stratification (for Cd, Mn, and Zn). In the Ottawa River, the main tributary, oxyhydroxide formation in summer governs seasonal patterns of Al, Fe, Mn, Cd, Co and Zn. The downstream section of the St. Lawrence River is a transit zone in which seasonal variations are mainly driven by the mixing of the different water masses and the large input of suspended particulate matter from erosion. The budget of all dissolved elements, except Fe and Zn, was balanced, as the budget of particulate elements (except Cd and Zn). The main sources of metals to the St. Lawrence River are erosion and inputs from tributaries and Lake Ontario. Direct anthropogenic discharges into the river accounted for less than 5% of the load, except for Cd (10%) and Zn (21%). The fluxes in transfer of dissolved Cd, Co, Cu and Zn species from the river to the lower St. Lawrence estuary were equal to corresponding fluxes calculated for Quebec City since the distributions of dissolved concentrations of these metals versus salinity were conservative. For Fe, the curvature of the dilution line obtained suggests that dissolved species were removed during early mixing.

The impact of variably saturated conditions on hydrogeochemical changes during artificial recharge of groundwater by Janek Greskowiak; Henning Prommer; Gudrun Massmann; Colin D. Johnston; Gunnar Nützmann; Asaf Pekdeger (1409-1426).
Artificial recharge of groundwater is often used to either purify partially treated wastewater or to enhance the quality of surface water by percolation through a variably saturated zone. In many cases, the most substantial purification process within the infiltration water is the redox-dependent biodegradation of organic substances. The present study was aimed at understanding the spatial and temporal distribution of the redox reactions that develop below an artificial recharge pond near Lake Tegel, Germany. At this site, like at many artificial recharge sites, the hydraulic regime immediately below the pond is characterised by cyclic changes between saturated and unsaturated conditions. These changes, which occur during each operational cycle, result from the repeated formation of a clogging layer at the pond bottom. Regular hydrogeochemical analyses of groundwater and seepage water in combination with continuous hydraulic measurements indicate that NO 3 - - and Mn-reducing conditions dominate beneath the pond as long as water-saturated conditions prevail. Manganese-, Fe- and SO 4 2 - -reducing conditions are confined to a narrow zone directly below the clogging layer and in zones of lower hydraulic conductivity. The formation of the clogging layer leads to a steady decrease of the infiltration rate, which ultimatively causes a shift to unsaturated conditions below the clogging layer. Atmospheric O2 then starts to penetrate from the pond fringes into this region, leading to: (i) the re-oxidation of the previously formed sulphide minerals and (ii) the enhanced mineralisation of sedimentary particulate organic C. The mineralisation of sedimentary particulate organic C leads to an increased H2CO3 production and subsequent dissolution of calcite.

Origin of sulfur rich oils and H2S in Tertiary lacustrine sections of the Jinxian Sag, Bohai Bay Basin, China by Chunfang Cai; Richard H. Worden; George A. Wolff; Simon Bottrell; Donglian Wang; Xin Li (1427-1444).
Very high S oils (up to 14.7%) with H2S contents of up to 92% in the associated gas have been found in the Tertiary in the Jinxian Sag, Bohai Bay Basin, PR China. Several oil samples were analyzed for C and S stable isotopes and biomarkers to try to understand the origin of these unusual oil samples.The high S oils occur in relatively shallow reservoirs in the northern part of the Jinxian Sag in anhydrite-rich reservoirs, and are characteristic of oils derived from S-rich source rocks deposited in an enclosed and productive stratified hypersaline water body. In contrast, low S oils (as low as 0.03%) in the southern part of the Jinxian Sag occur in Tertiary lacustrine reservoirs with minimal anhydrite. These southern oils were probably derived from less S-rich source rocks deposited under a relatively open and freshwater to brackish lake environment that had larger amounts of higher plant inputs.The extremely high S oil samples (>10%) underwent biodegradation of normal alkanes resulting in a degree of concentration of S in the residual petroleum, although isoprenoid alkanes remain showing that biodegradation was not extreme. Interestingly, the high S oils occur in H2S-rich reservoirs (H2S up to 92% by volume) where the H2S was derived from bacterial SO4 reduction, most likely in the source rock prior to migration. Three oils in the Jinxian Sag have δ 34S values from +0.3‰ to +16.2‰ and the oil with the highest S content shows the lightest δ 34S value. This δ 34S value for that oil is close to the δ 34S value for H2S (∼0‰). It is possible that H2S was incorporated into functionalized compounds within the residual petroleum during biodegradation at depth in the reservoir thus accounting for the very high concentrations of S in petroleum.