Katherine Walton-Day

Assessment of metal loads in watersheds affected by acid mine drainage by using tracer injection and synoptic sampling: Cement Creek, colorado, USA

Watersheds in mineralized zones may contain many mines, each of which can contribute to acidity and the metal load of a stream. In this study the authors delineate hydrogeologic characteristics determining the transport of metals from the watershed to the stream in the watershed of Cement Creek, Colorado. Combining the injection of a chemical tracer, to determine a discharge, with synoptic sampling, to obtain chemistry of major ions and metals, spatially detailed load profiles are quantified. Using the discharge and load profiles, the authors (1) identified sampled inflow sources which emanate from undisturbed as well as previously mined areas; (2) demonstrate, based on simple hydrologic balance, that unsampled, likely dispersed subsurface, inflows are significant; and (3) estimate attenuation. For example, along the 12-km study reach, 108 kg per day of Zn were added to Cement Creek. Almost half of this load came from 10 well-defined areas that included both mined and non-mined parts of the watershed. However, the combined effect of many smaller inflows also contributed a substantial load that could limit the effectiveness of remediation. Of the total Zn load, 58.3 kg/day came from stream segments with no visible inflow, indicating the importance of contributions from dispersed subsurface inflow. The subsurface inflow mostly occurred in areas with substantial fracturing of the bedrock or in areas downstream from tributaries with large alluvial fans. Despite a pH generally less than 4.5, there was 58.4 kg/day of Zn attenuation that occurred in mixing zones downstream from inflows with high pH. Mixing zones can have local areas of pH that are high enough for sorption and precipitation reactions to have an effect. Principal component analysis classified inflows into 7 groups with distinct chemical signatures that represent water-rock interaction with different mineral-alteration suites in the watershed. The present approach provides a detailed snapshot of metal load for the watershed to support remediation decisions, and quantifies processes affecting metal transport.

The Interaction of Natural Organic Matter with Iron in a Wetland (Tennessee Park, Colorado) Receiving Acid Mine Drainage

Pore water from a wetland receiving acid mine drainage was studied for its iron and natural organic matter (NOM) geochemistry on three different sampling dates during summer 1994. Samples were obtained using a new sampling technique that is based on screened pipes of varying length (several centimeters), into which dialysis vessels can be placed and that can be screwed together to allow for vertical pore-water sampling. The iron concentration increased with time (through the summer) and had distinct peaks in the subsurface. Iron was mainly in the ferrous form; however, close to the surface, significant amounts of ferric iron (up to 40% of 2 mmol L-1 total iron concentration) were observed. In all samples studied, iron was strongly associated with NOM. Results from laboratory experiments indicate that the NOM stabilizes the ferric iron as small iron oxide colloids (able to pass a 0.45μm dialysis membrane). We hypothesize that, in the pore water of the wetland, the high NOM concentrations (>100 mg C L-1) allow formation of such colloids at the redoxcline close to the surface and at the contact zone to the adjacent oxic aquifer. Therefore, particle transport along flow paths and resultant export of ferric iron from the wetland into ground water might be possible.

Evaluating remedial alternatives for an acid mine drainage stream: a model post audit.

A post audit for a reactive transport model used to evaluate acid mine drainage treatment systems is presented herein. The post audit is based on a paired synoptic approach in which hydrogeochemical data are collected at low (existing conditions) and elevated (following treatment) pH. Data obtained under existing, low-pH conditions are used for calibration, and the resultant model is used to predict metal concentrations observed following treatment. Predictions for Al, As, Fe, H(+), and Pb accurately reproduce the observed reduction in dissolved concentrations afforded by the treatment system, and the information provided in regard to standard attainment is also accurate (predictions correctly indicate attainment or nonattainment of water quality standards for 19 of 25 cases). Errors associated with Cd, Cu, and Zn are attributed to misspecification of sorbent mass (precipitated Fe). In addition to these specific results, the post audit provides insight in regard to calibration and sensitivity analysis that is contrary to conventional wisdom. Steps taken during the calibration process to improve simulations of As sorption were ultimately detrimental to the predictive results, for example, and the sensitivity analysis failed to bracket observed metal concentrations.

New light on a dark subject: On the use of fluorescence data to deduce redox states of natural organic matter (NOM)

Abstract.This paper reports the use of excitation-emission matrix fluorescence spectroscopy (EEMS), parallel factor statistical analysis (PARAFAC), and oxidation-reduction experiments to examine the effect of redox conditions on PARAFAC model results for aqueous samples rich in natural organic matter. Fifty-four aqueous samples from 11 different geographic locations and two plant extracts were analyzed untreated and after chemical treatments or irradiation were used in attempts to change the redox status of the natural organic matter. The EEMS spectra were generated and modeled using a PARAFAC package developed by Cory and McKnight (2005). The PARAFAC model output was examined for consistency with previously reported relations and with changes expected to occur upon experimental oxidation and reduction of aqueous samples. Results indicate the implied fraction of total sample fluorescence attributed to quinone-like moieties was consistent (0.64 to 0.78) and greater than that observed by Cory and McKnight (2005). The fraction of the quinone-like moieties that was reduced (the reducing index, RI) showed relatively little variation (0.46 to 0.71) despite attempts to alter the redox status of the natural organic matter. The RI changed little after reducing samples using zinc metal, oxidizing at high pH with air, or irradiating with a Xenon lamp. Our results, however, are consistent with the correlations between the fluorescence indices (FI) of samples and the ratio of PARAFAC fitting parameters suggested by Cory and McKnight (2005), though we used samples with a much narrower range of FI values.

Mechanisms controlling Cu, Fe, Mn, and Co profiles in peat of the Filson Creek Fen, northeastern Minnesota

Abstract Filson Creek Fen, located in northeastern Minnesota, overlies a Cu-Ni sulfide deposit. A site in the fen was studied to evaluate the hydrogeochemical mechanisms governing the development of Fe, Mn, Co, and Cu profiles in the peat. At the study site, surface peat approximately 1 m thick is separated from the underlying mineralized bedrock by a 6–12 m thickness of lake and glaciofluvial sediments and till. Concentrations of Fe, Mn, Co, and Cu in peat and major elements in pore water delineate a shallow, relatively oxidized, Cu-rich zone overlying a deeper, reduced, Fe-, Mn-, and Co-rich zone within the peat. Sequential metal extractions from peat samples reveal that 40–55% of the Cu in the shallow zone is associated with organic material, whereas the remaining Cu is distributed between iron-oxide, sulfide, and residual fractions. Sixty to seventy percent of the Fe, Mn, and Co concentrated in the deeper zone occur in the residual phase. The metal profiles and associations probably result from non-steady-state input of metals and detritus into the fen during formation of the peat column. The enrichment of organic-associated Cu in the upper, oxidized zone represents a combination of Cu transported into the fen with detrital plant fragments and soluble Cu, derived from weathering of outcrop and subcrop of the mineral deposit, transported into the fen, and fixed onto organic matter in the peat. The variable stratigraphy of the peat indicates that weathering processes and surface vegetation have changed through time in the fen. The Fe, Mn, and Co maxima at the base of the peat are associated with a maximum in detrital matter content of the peat resulting from a transition between the underlying inorganic sedimentary environment to an organic sedimentary environment. The chemistry of sediments and ground water collected beneath the peat indicate that mobilization of metals from sulfide minerals in the buried mineral deposit or glacial deposits is minimal. Therefore, the primary source of Cu to the peat at the study site is outcrops and shallow subcrops of the mineral deposit adjacent to the fen.

The precipitation of indium at elevated pH in a stream influenced by acid mine drainage

Indium is an increasingly important metal in semiconductors and electronics and has uses in important energy technologies such as photovoltaic cells and light-emitting diodes (LEDs). One significant flux of indium to the environment is from lead, zinc, copper, and tin mining and smelting, but little is known about its aqueous behavior after it is mobilized. In this study, we use Mineral Creek, a headwater stream in southwestern Colorado severely affected by heavy metal contamination as a result of acid mine drainage, as a natural laboratory to study the aqueous behavior of indium. At the existing pH of ~3, indium concentrations are 6-29μg/L (10,000× those found in natural rivers), and are completely filterable through a 0.45μm filter. During a pH modification experiment, the pH of the system was raised to >8, and >99% of the indium became associated with the suspended solid phase (i.e. does not pass through a 0.45μm filter). To determine the mechanism of removal of indium from the filterable and likely primarily dissolved phase, we conducted laboratory experiments to determine an upper bound for a sorption constant to iron oxides, and used this, along with other published thermodynamic constants, to model the partitioning of indium in Mineral Creek. Modeling results suggest that the removal of indium from the filterable phase is consistent with precipitation of indium hydroxide from a dissolved phase. This work demonstrates that nonferrous mining processes can be a significant source of indium to the environment, and provides critical information about the aqueous behavior of indium.

Synoptic sampling and principal components analysis to identify sources of water and metals to an acid mine drainage stream

Combining the synoptic mass balance approach with principal components analysis (PCA) can be an effective method for discretising the chemistry of inflows and source areas in watersheds where contamination is diffuse in nature and/or complicated by groundwater interactions. This paper presents a field-scale study in which synoptic sampling and PCA are employed in a mineralized watershed (Lion Creek, Colorado, USA) under low flow conditions to (i) quantify the impacts of mining activity on stream water quality; (ii) quantify the spatial pattern of constituent loading; and (iii) identify inflow sources most responsible for observed changes in stream chemistry and constituent loading. Several of the constituents investigated (Al, Cd, Cu, Fe, Mn, Zn) fail to meet chronic aquatic life standards along most of the study reach. The spatial pattern of constituent loading suggests four primary sources of contamination under low flow conditions. Three of these sources are associated with acidic (pH <3.1) seeps that enter along the left bank of Lion Creek. Investigation of inflow water (trace metal and major ion) chemistry using PCA suggests a hydraulic connection between many of the left bank inflows and mine water in the Minnesota Mine shaft located to the north-east of the river channel. In addition, water chemistry data during a rainfall-runoff event suggests the spatial pattern of constituent loading may be modified during rainfall due to dissolution of efflorescent salts or erosion of streamside tailings. These data point to the complexity of contaminant mobilisation processes and constituent loading in mining-affected watersheds but the combined synoptic sampling and PCA approach enables a conceptual model of contaminant dynamics to be developed to inform remediation.

Principal locations of major-ion, trace-element, nitrate, and Escherichia coli loading to Emigration Creek, Salt Lake County, Utah, October 2005

.......................................................................................................................................................