Pathways of Metal Transfer from Mineralized Sources to Bioreceptors: A Synthesis of the Mineral Resources Program’s Past Environmental Studies in the and Future Research Directions

Chapter 6—Impacts of historical mining in the Coeur d’Alene River Basin By Laurie S. Balistrieri1, Stephen E. Box1, Arthur A. Bookstrom1, Robert L. Hooper2, and J. Brian Mahoney2

CONTENTS BACKGROUND ore shoots in Precambrian rocks of the Background 1 Mining began in the late 1880s in the Belt Supergroup (Fryklund, 1964; Discussion of mining impacts 3 Coeur d’Alene mining district in north- Hobbs and others, 1965; Zartman and What is known about the distributions, con- ern (fig. 1). Although only two Stacey, 1971; Bennett and centrations, and speciation of Pb and Zn in the Coeur d’Alene River basin? 3 mines, the Galena and Lucky Friday, Venkatakrishnan, 1982; Leach and Particulate lead. 3 currently are operating, more than 90 others, 1988; Criss and Fleck, 1990). Dissolved zinc 4 historical mines exist in this region The veins are separated into two major Solid-phase speciation 5 (Bennett and others, 1989). Most of the types by ore mineralogy: (1) lead- and What is known about the processes that mobi- mines are along the South Fork of the zinc-rich veins having argentiferous lize Pb and Zn from their sources and Coeur d’Alene River and its major galena (PbS) and sphalerite (ZnS) and then act to physically and biogeochemically redistribute them? 7 tributaries (Bennett and others, 1989). (2) silver-rich veins having argentifer- Physical processes 7 Total production records indicate that ous tetrahedrite [(Cu, Ag)10(Fe, Biogeochemical processes 8 this district ranks among supergiants Zn)2(As, Sb)4S13] and minor galena and Modeling of coupled physical and bio- (top 1 percent of world producers) for sphalerite. The vein types are spatially geochemical processes 10 silver (34,300 metric tons Ag) and lead separated, may represent one or two What is known about impacts of Pb and Zn to (7,290,000 metric tons Pb) and among ages of mineralization, and were depos- biota, and can specific processes that in- fluence bioavailability be identified? 12 giants (top 10 percent of world produc- ited from fluids of metamorphic origin What are the major gaps in knowledge re- ers) for zinc (2,870,000 metric tons Zn) (Leach and others, 1998; Long, 1998a).

maining from this research? 13 (Long, 1998a, 1998b). Pyrite (FeS2) is ubiquitous but variable Acknowledgments 14 Ore deposits in the district are in abundance in the veins. Most veins References 14 steeply dipping quartz veins and sider- contain small amounts of chalcopyrite ite (FeCO ) veins containing (CuFeS ). Minor minerals include 1United States Geological Survey 3 2 2University of Wisconsin Eau Claire stratigraphically controlled Pb-Zn-Ag arsenopyrite (FeAsS) and pyrrhotite

1 Chapter 6 (Fe1-xS). Host rocks are primarily directly into the Coeur d’Alene River 1917 and 1982 (Bennett and others, quartzite and argillite, which contain and its tributaries before environmental 1989). This smelter released more than some interbedded carbonate-bearing regulations required the installation of 3,300 metric tons of Pb to the atmo- rocks. Studies at the Lucky Friday tailings ponds in 1968. A preliminary sphere between 1965 and 1981. An Mine indicate that wall rocks around accounting by Long (1998b) has esti- area of 54 km2 surrounding the smelter veins are altered and typically contain mated that 56 million metric tons of was listed by the U.S. Environmental 10 to 15 percent carbonate minerals. tailings containing 2,200 metric tons of Protection Agency (EPA) as one of the Concentric zonation with respect to Ag, 800,000 metric tons of Pb, and at nation’s largest Superfund sites in three carbonate minerals [siderite least 650,000 metric tons of Zn were 1983 (U.S. Environmental Protection

(FeCO3), ankerite [CaFe(CO3)2], and dumped into the river system. Stream Agency, 1994). The listing was calcite (CaCO3)] is common in the transport, especially during major flood prompted, in part, by very high levels altered wall rocks (Gitlin, 1986). The events, has redistributed and continues of Pb in the blood of children living in predominant gangue minerals are to redistribute metal-enriched sediment Kellogg. A Natural Resource Damages siderite and quartz. The absolute and from its sources for distances of more (NRD) lawsuit filed by the Federal relative abundances of sulfide and than 240 km downstream throughout government and two Native American gangue minerals vary significantly the channel of the South Fork and main tribes is awaiting a decision after 2 between different vein systems. stem of the Coeur d’Alene River and months of court testimony in 2001. Milling and ore concentration prac- their floodplains, into Lake Coeur The EPA completed a 3-year Remedial tices varied over time in the district d’Alene, and into the Spokane River Investigation/Feasibility Study (RI/FS) because of changing technology and (Horowitz and others, 1993; Horowitz for the entire Coeur d’Alene Basin in economics. Early ore separation meth- and others, 1995; Bookstrom and 2001. The proposed plan, in its public ods, which included coarse crushing others, 2001; Box and others, 2001). comment period as of May 2002, and gravity (“jig”) mineral separation Because the river gradient is steeper recommends a 30-year, $300 million methods, were not very efficient. Jig and the flow faster in the upper Coeur remedial program as the first incre- tailings produced before 1915 ranged d’Alene River system, the major reposi- ment of a longer remediation schedule. from coarse gravel to fine powder and tories of the discharged mine tailings EPA is expected to issue a Record of were still very rich in metals. Develop- were the channel and floodplain of the Decision laying out its remediation ment of more efficient flotation meth- lower Coeur d’Alene River (that is, plan in mid-2002. The State of Idaho ods between 1915 and 1925 resulted in between Cataldo and Harrison) and has completed some remediation tailings with finer grain size (fine sand Lake Coeur d’Alene. projects and is currently doing several and finer) and lower metal concentra- The Bunker Hill lead smelter in more within the district but outside of tions. Most tailings were deposited Kellogg (see fig. 1) operated between the Superfund site.

2 Chapter 6 Although our data encompass a The EPA and U.S. Fish and Wildlife cause of closure of the smelters and wide range of elements, the following Service provided supplemental funding. environmental regulations that prohibit discussion of our work focuses on the the dumping of tailings into rivers. behavior of Pb and Zn in the near- DISCUSSION OF MINING However, historically produced particu- surface environment because of their IMPACTS late Pb from smelter fallout and mill importance to environmental and tailings is constantly being redistributed health issues in the Coeur d’Alene What is known about the distribu- by wind and water. Because Superfund River basin. As discussed in the intro- tions, concentrations, and specia- work focused on cleaning up smelter ductory chapter (ch. 1), we examine tion of Pb and Zn in the Coeur fallout in houses and yards, the biggest the results of our work in light of four d’Alene River basin? remaining challenge is grappling with the questions: environmental and health impacts of 1. What is known about the distribu- Particulate Lead fluvially distributed tailings. tions, concentrations, and specia- Pb exists primarily in the particulate or The origin of the fluvially deposited tion of Pb and Zn in this river solid phase rather than the dissolved tailings is the historical milling and basin? phase. This is because of the low solubil- processing of ore within the district and 2. What is known about the processes ity of Pb minerals and the high affinity of subsequent disposal of tailings into the that release Pb and Zn from their dissolved Pb for metal oxide particles at river system. The concentration of sources and then act to physically neutral pH. Health problems in the basin metals in tailings from mills in the and biogeochemically redistribute for humans and wildlife are linked to Coeur d’Alene mining district decreased them? high concentrations of particulate Pb in irregularly through time as ore concen- 3. What is known about the impacts surface soils and sediment and to inges- tration methods improved (fig. 2). of Pb and Zn on biota and can we tion of those particles. Elevated concen- Gravity separation or jigging was the ore identify specific processes that trations of particulate Pb are associated concentration method used between influence bioavailability? with soils that formed over mineralized 1885 and 1925. Data from the historical 4. What additional work needs to be rocks in the area (Gott and Cathrall, Morning Mill, the remnants of which are done to provide a more thorough 1980), tailings from mills that processed just west of Mullan, indicates that understanding of Pb and Zn cy- the mineralized rock (Long, 1998b), and tailings contained between 4 and 9 cling in this basin? atmospheric fallout from smelters that percent Pb and Zn during this time (fig. Our work was done as part of the operated in the mining district (U.S. 2). Flotation methods, used between Coeur d’Alene Project that was funded Environmental Protection Agency, 1994). 1925 and 1968, produced tailings con- primarily by the Mineral Resources There are no new sources of particulate taining lower metal concentrations (<1.5 Program of the U.S. Geological Survey. Pb from smelters or tailings today be- percent Pb and Zn) (fig. 2).

3 Chapter 6 The distribution of Pb as a function of diminution of sedimentation rate or Box and others, 2001). These calculations depth in the sediments of the Coeur metal content in river bank or Lake indicate that there are 200,000±100,000 d’Alene River valley reflects the onset of Coeur d’Alene sediments has been noted metric tons of Pb in the South Fork drain- mining and history of milling practices since the dumping of tailings dumping age basin, 250,000±62,000 metric tons in within the district (Bookstrom and into the river was stopped in 1968 the lower Coeur d’Alene River valley others, 2001; Box and others, 2001). (Horowitz and others, 1993; Box and between Cataldo and Harrison, and Before mining began in the late 1880s, others, 2001). 292,000±145,000 metric tons in Lake the concentration of Pb was <30 ppm in In sediments finer than 175 mm in the Coeur d’Alene. Estimates have not yet the silty clay and fine sand of the basin bed of the Spokane River, the Pb content been made for the North Fork drainage (Bookstrom and others, 2001). Similar ranges from background concentrations basin or the Spokane River. About 99 concentrations are observed today in the (<30 ppm) to about 2,500 ppm, whereas percent of the Pb contained in these deepest portion of cores collected from Zn concentrations vary from about 50 to sediments was added as a result of mining- transects across the river channel at 5,000 ppm (fig. 4). A mixing model related activities. Of the Pb in the lower Killarney (between Rose Lake and suggests these sediments are composed Coeur d’Alene River valley, approximately Harrison) and from the adjacent flood- of three types of material that have 59 percent is in the floodplain, 37 percent plain (fig. 3). Changes in milling prac- different sources: (1) background mate- in the river channel, and 4 percent in the tices over time also are reflected in the rial with low Pb (20 ppm) and low Zn riverbanks. The estimated sum of Pb added varying concentrations of Pb with depth (50 ppm) concentrations, (2) material, to sediments of the South Fork River across this transect. Sediments mixed possibly an authigenic flocculent, that is valley, Coeur d’Alene River valley, and with early jig tailings that were depos- poor in Pb (50 ppm) and rich in Zn Lake Coeur d’Alene (739,000 + 305,000 ited before about 1925 have the highest (5,000 ppm), and (3) material derived metric tons) amounts to about 87 percent concentrations of Pb (up to 36,000 ppm). from the Coeur d’Alene River that has of the estimated Pb (850,000 + 10,000 A sharp drop in Pb concentration to less moderately high concentrations of both metric tons) that was dumped into tribu- than 10,000 ppm, followed by a gradual Pb (2,500 ppm) and Zn (4,000 ppm) tary streams by the mills in the mining upward decrease in Pb contents to 3,000- (Box and Wallis, 2002). district. 5,000 ppm, mark sediments from the era Calculations of the volume of sediment Dissolved Zinc of flotation methods. A gradual upward and associated Pb in various areas of the coarsening trend from silt at the base to river basin were recently done using digital Although large amounts of particulate medium sand at the top is typical of the maps of the South Fork and lower Coeur Zn reside in the Coeur d’Alene River river channel and bank deposits. Deposi- d’Alene River and chemical analyses of basin, it is the dissolved form of Zn that tion rates are still high on the river banks many sediment samples (Bookstrom and governs water quality in this region (average of 0.8 cm/yr) and little, if any, others, 1999; Bookstrom and others, 2001; because of its impact on the health of

4 Chapter 6 biota, particularly fish. Because of this seeps, and river water) affected by ore except from the Kellogg Tunnel, has impact, we compare concentrations of deposits or mining wastes within the slightly alkaline pH values (median pH dissolved Zn observed in the basin to the Coeur d’Alene River basin. Both pH and = 7.34) and moderately high dissolved Criterion Continuous Concentration dissolved concentrations of Zn show Zn concentrations (up to 58 mg/L). (CCC) for Zn, which is the highest large variations, and many samples, Water from rivers is mostly near neutral concentration of total recoverable Zn to particularly of ground water, do not meet and tends to have the lowest dissolved which aquatic life can be exposed for an water quality criteria (Mink and others, Zn concentrations, although the median extended period of time (4 days) without 1971; U.S. Geological Survey, 1973; value is still above the CCC. deleterious effects (U.S. General Ser- McCulley Frick and Gilman, 1994; Solid-phase speciation vices Administration, 1999). Within the Balistrieri and others, 1998; Golder basin, total recoverable Zn in water Associates, 1998; TerraGraphics Envi- We have conducted two studies on the samples typically is equal to dissolved ronmental Engineering, 1998; Balistrieri speciation of metal-enriched particles in Zn (Brennan and others, 2000; Brennan and others, 2000; McCulley Frick and the Coeur d’Alene River valley. The and others, 2001). The CCC for Zn in Gilman, 2000). Values of pH range from intent was to look at changes in mineral- Idaho is a function of the hardness of the a low of 2.72 at the Kellogg Tunnel (adit ogy and geochemical availability of water (U.S. General Services Adminis- drainage) to a high of 9.1 in the Coeur particulate Pb and Zn as they are dis- tration, 1999). For example, for a range d’Alene River. Dissolved Zn concentra- persed from their primary sources as in hardness of 15 to 65 mg/L, as ob- tions range from 1.2 mg/L to 759 mg/L, sulfide minerals (galena and sphalerite) served in the Coeur d’Alene River, the and vary significantly at near-neutral pH in the ore deposits and to determine how corresponding CCC for Zn ranges from values. Table 1 summarizes the range the redox characteristics of the deposi- 21 to 73 mg/L (U.S. General Services and median values for pH and dissolved tional environment influence the miner- Administration, 1999). A water quality Zn concentrations in the various waters. alogy and geoavailability of Pb and Zn. criterion for pH has also been set and Except for water from the Kellogg Our conclusions are based on mineral- ranges from 6.5 to 9 for freshwater (U.S. Tunnel, ground water under tailings- ogical and geochemical leach studies of Environmental Protection Agency, bearing floodplain sediments and tail- the particles. 1999). ings piles has the lowest pH values and The first study looked at the specia- Dissolved Zn concentrations are highest dissolved Zn concentrations. tion of Pb in particles from paired sites plotted as a function of pH in figure 5 Porewater in metal-contaminated sedi- at three locations within the mining for different types of water (adit drain- ment mostly tends to be slightly acidic district. At each location, the paired sites age, ground water from deep wells, (median pH = 6.57) and can have moder- are the floodplain, which contains porewater in the upper 30 cm of water- ately high dissolved Zn concentrations tailings from the jig era, and the saturated sediments, water from tailings (up to 70 mg/L). Most adit drainage, nearby river channel. Both types of

5 Chapter 6 site are oxidizing environments. ated with Fe and Mn oxides within a Periodic flooding transports pri- Mineralogical (R.L. Hooper and C. short distance (<10 km) from the ore mary ore minerals (sulfides) and Rowe, unpublished data, 2000) and deposits. These results are consistent secondary mineral phases (oxides and leach data using the methods of with observations of mineralogical carbonates) from upstream tailings Gasser (Gasser and others, 1996) for changes in Pb downstream from lead- deposits into the lower river valley. these particles are shown in table 2. zinc-fluorite-barite deposits in north- Preliminary results of our analyses of The Pb, Fe, and Mn contents of these east England (Hudson-Edwards and particles in the lower river valley samples vary substantially, consider- others, 1996). suggest that their mineralogy is very ably higher concentrations generally The second study looked at the complex and that they contain consid- occurring in sediment from the flood- speciation of Pb and Zn in particles as a erable amounts of amorphous and plain. The mineralogical data indicate function of depth from four different nanocrystalline material. The fate of the importance of Fe and Mn oxides environments (river bed, levee banks, particles, which may involve mineral as host phases for Pb (also see fig. 6) wetlands, and lateral lakes) in the lower dissolution, metal mobilization and and, in some cases, the formation of Coeur d’Alene River valley using a migration, and mineral precipitation, substantial amounts of Pb carbonates combination of mineralogical and depends on the redox characteristics, and sulfates. Trace amounts of chemical techniques (SEM, TEM, and permeability, organic content, and sphalerite are present in all samples. sequential leaches) (Hooper and microbial activity of the depositional The sequential extraction data are in Mahoney, 2000, 2001). Care was taken environment. Particles in the bed of accord with the mineralogical data in to maintain the redox characteristics of the river below the sediment-water that they indicate large fractions of Pb the samples during collection, storage, interface are in a reducing environ- associated with Fe and Mn oxides and analyses. A sequential extraction ment, which acts as a sink for detrital

(that is, in the EDTA or HNO3 leach) method (Tessier and others, 1979) was sulfides and carbonates (fig. 8). or, in the case of sample 205, angles- calibrated by determining the mineral- Authigenic Fe, Pb, and Zn sulfides ite (that is, in the MgCl2 leach). ogy after each extraction step (fig. 7). appear to form near the sediment- Results from the leach simulating The individual steps in the sequential water interface. Because of water- gastric conditions suggest that a extraction scheme generally are not level changes throughout the year, major fraction of Pb in most of these specific for given mineral phases. particles in the levee environment, samples would be mobilized in the However, patterns of metal release, near the edge of the river, cycle human stomach. This work indicates considering all of the extraction steps between dry and wet conditions. This that the speciation of particulate Pb along with the calibration results, results in oxidizing zones containing changes from a sulfide mineral to Pb provide information about the domi- oxide-coated grains interspersed with carbonates and sulfates or Pb associ- nant phases in the samples. reducing zones dominated by detrital

6 Chapter 6 and authigenic carbonate and sulfide producing a nanocrystalline biofilm on duced by jigging and flotation methods phases. Unexpectedly, detrital sphaler- particles that is characterized by ZnS within the district. About 51 percent of ite is present in oxidized levee samples, and nonstoichiometric PbS, FeS, and those were disposed of in the Coeur indicating that some fraction of the mixed metal sulfides. d’Alene River and its tributaries, 37 total amount of the mineral is resistant percent were placed in impoundments or to oxidation. Sediment at the top of What is known about the pro- used as stope fill, and 12 percent were levees resides in oxidizing environ- cesses that mobilize Pb and Zn stockpiled along the floodplain. Hydro- ments, and samples from these environ- from their sources and then act to logic transport, especially during flood ments, contain small amounts of detrital physically and biogeochemically events, resulted in sorting of this metal- galena (PbS) and greater amounts of redistribute them? enriched material and in the deposition of cerrusite (PbCO ) and Pb-Fe-Mn oxides fluvial tailings deposits not only within 3 Physical processes (fig. 8). In some samples Pb is associ- the mining district, but also throughout ated with Mn oxides, and in others it is Several physical processes play impor- the lower Coeur d’Alene River valley, in associated with Fe oxides. Zn exists tant roles in determining the distribution Lake Coeur d’Alene, and in the Spokane primarily with Fe-Mn oxides. The and concentration of Pb and Zn in the River. There is a potential long-term wetland environment is anoxic, and Coeur d’Alene River basin, including source of metal-enriched material to the most of the particulate Pb and Zn exist those involved in historical mining floodplain, Lake Coeur d’Alene, and the as microcrystalline and nanocrystalline practices; hydrologic transport and Spokane River because a large fraction of to amorphous authigenic sulfides seasonal fluctuations in water levels; Pb (about 41 percent of Pb in the lower associated with biofilms, rootlets, and sorting of particles by size fractions Coeur d’Alene River valley) resides or is bacteria. Anoxic conditions also exist in during flood events; mixing of different stored in the river channel and riverbanks the sediments of the lateral lakes. source waters; and transport of dissolved (Box and others, 2001) and is subject to Particles in the lateral lakes contain metals across the sediment-water inter- movement during high flows. nanocrystalline inorganic and biogenic face by molecular diffusion. Each of Pb and Zn data for particles in the sulfides in the upper third of the metal- these is discussed here. river bed and on the levee banks (Box enriched sediment and increasing The milling practices and subsequent and others, 2001) and for particles amounts of silt-size detrital sulfides disposal of tailings into the river system suspended in the water of the Coeur (especially sphalerite) closer to the before 1968 were important factors in d’Alene River during two flood events premining surface. In both the wetland determining the present day geochemical (Box and others, in press) are used to and lateral lake environments, micro- characteristics of the basin. Long (1998b) illustrate the importance of sources of bial activity is extremely effective in estimated that 109 million metric tons of sediment to the river (fig. 9). removing metals from the water and metal-enriched mill tailings were pro- Riverbanks and natural levees of the

7 Chapter 6 Coeur d’Alene River and alluvial Another physical process that influ- into the water column of the lake are terraces of the South Fork have lower ences the distribution of elements is similar to fluxes supplied to the lake by Zn than Pb concentrations because Zn mixing of different source waters. Adit the St. Joe and Coeur d’Alene Rivers. is preferentially leached from sedi- drainage within the district that is en- ments stored in oxidizing environ- riched in dissolved elements mixes with Biogeochemical processes ments. In contrast, bed sediments in the rivers and creeks that have lower concen- river channel are more enriched in Zn trations of elements (Balistrieri and Water-rock interactions act to mobilize than Pb. The major rain-on-snow flood others, 1998). The North Fork of the and redistribute elements within the river of 1996 began suddenly, when Lake Coeur d’Alene River supplies less metal- basin. Microorganisms mediate many of Coeur d’Alene was low and hydraulic enriched water and sediment to the lower these reactions. Oxidation reactions head between the river and the lake was Coeur d’Alene River than the South Fork involving primary sulfide minerals in the high. Floodwaters ran red with oxidized at their confluence above Cataldo. Also, ore deposits result in the release of suspended sediment, and Zn/Pb ratios of variations in metal loading along the metals, sulfate, and in some cases (for suspended sediment were less than one, South Fork and its tributaries suggest example, pyrite oxidation) acid to the indicating that most of the suspended mixing of surface water with more metal- water. This acidity can be neutralized by sediment was mobilized from oxidizing enriched ground water (Barton, 2002). the dissolution of buffering minerals, environments. By contrast, the major Dissolved Zn, other metals, and nutri- primarily carbonates such as calcite or 1997 spring runoff flood began gradu- ents can be supplied to Lake Coeur ankerite. The relative molar amounts of ally and continued for about a month, d’Alene by inflowing rivers and by pyrite and carbonate minerals in the with relatively clear waters in the upper transport across the sediment-water deposits and host rocks that react or the basin, sluggish currents along the Coeur interface (benthic flux). Benthic fluxes reacting pyrite-to-calcite ratio can be d’Alene River, and floodwater backing were determined from concentrations in predicted from the composition of drain- up from Lake Coeur d’Alene as it the water at the bottom of the lake and in age from adits (Balistrieri and others, overfilled. During this 1997 and previ- porewater in the sediment using Fick’s 1999; Balistrieri and others, 2002). The ous spring runoff floods, Zn/Pb ratios of First Law and time-series data from in- predicted reacting pyrite-to-calcite ratios suspended sediment were greater than situ benthic flux chambers (Balistrieri, (assuming precipitation of ferrihydrite) one, indicating that fines, winnowed out 1998; Kuwabara and others, 2000). The for deposits in the Coeur d’Alene mining of riverbed sediment, predominated over results indicate that transport of dissolved district range from 0.02 to 0.5 for near- sediment from oxidizing environments. elements and species is controlled by neutral drainage to about 1 for the acidic In both events, however, sandy molecular diffusion and that the magni- drainage from the Kellogg Tunnel at the riverbank deposits were derived from tudes of the benthic fluxes of dissolved Bunker Hill mine site (fig. 10). Data for mobilization of nearby bed sediments. Zn, Cd, phosphate, and nitrogen species drainage from other polymetallic vein

8 Chapter 6 deposits in the Colorado Mineral Belt and porewater and assessed the ability of (transported scenario) rather than being the Humboldt River Basin, Nevada, are thermodynamic models to predict the separated from the water and remaining also shown in figure 10 for comparison precipitation of mineral phases and in the aquifer (retained scenario) along (Plumlee and others, 1993; Plumlee and adsorption of metals onto metal oxide a flow path. Because pH increases others, 1999; Nash, 2000). Theoretically, phases (Paulson and Balistrieri, 1999; along the flow path, greater metal neutralization of acid generated during Tonkin and others, 2002) using removal occurs in the transported than oxidation of one mole of pyrite requires PHREEQC (Parkhurst and Appelo, 1999) in the retained case because the zone of between 2 and 4 moles of calcite (that is, and a database containing sorption precipitation of Fe oxide, which occurs reacting pyrite-to-calcite ratios between parameters for elements onto hydrous at a lower pH, is not separated from the 0.25 and 0.5) (Cravotta and others, 1990). ferric oxide (Dzombak and Morel, 1990). zone of adsorption of metals onto Fe Data for water draining polymetallic vein Many important conclusions resulted oxide, which occurs at a higher pH. deposits in Idaho, Colorado, and Nevada from this work. First, thermodynamic The influence of organic matter are consistent with the theoretical predic- modeling successfully predicted observed diagenesis on redox state has been tions, showing drainage pH values that pH changes, precipitation of ferrihydrite, demonstrated in porewater studies in the range from near neutral to alkaline when and removal of Cd, Cu, Pb, Zn, and, in sediments of Lake Coeur d’Alene and in reacting pyrite-to-calcite ratios for the some cases, As in the mixing experiments sediments along the river’s edge and in deposits are <0.3. (fig. 11). Second, the relative efficiency marshes of the lower Coeur d’Alene The precipitation of secondary miner- of removal of metals by adsorption onto River valley (Balistrieri, 1998; Balistrieri als such as carbonates (cerrusite and Fe oxide is Pb>Cu>> Zn~Cd. Third, Fe and others, 2000). Organic matter smithsonite), sulfates (Zn sulfate and oxide is a stronger adsorbent than Al decomposition occurs using a thermody- anglesite), and oxides (Fe and Mn oxide, and adsorbed organic matter is an namically predictable sequence of oxyhydroxides and Fe important adsorbent for Zn and Cd. oxidants (oxygen, nitrate, Mn oxide, Fe oxyhydroxysulfates) and adsorption of Fourth, the adsorption characteristics of oxide, and sulfate) and results in redox dissolved elements onto the secondary ferrihydrite and schwertmannite can be conditions that range from oxic to oxide phases occur after mobilization of described by the same set of adsorption suboxic, or to anoxic and sulfidic as species from the primary sulfide miner- parameters (that is, site density, surface depth increases in the sediment (Froelich als. The identity and composition of these area, and adsorption complexation and others, 1979). Redox state, in turn, precipitates were determined using constants). Fifth, more metal is re- influences the mobility of elements, mineralogical, chemical, and modeling moved by adsorption if Fe oxides that particularly Zn, because of the release of techniques. We conducted several sets of precipitate during the mixing of acidic associated elements during reduction of experiments that mixed surface water ground water and near-neutral surface oxide phases under suboxic conditions with ground water, adit drainage, or water remain suspended in the water and the insolubility of authigenic metal

9 Chapter 6 sulfide phases under anoxic and sulfidic and after remediation activities. One dissolved Zn, represented by the PZn term conditions. Porewater data from sedi- such model describing the behavior of in equation 1, act as an additional source ments in the lower Coeur d’Alene River dissolved Zn in Lake Coeur d’Alene is of dissolved Zn to the lake, whereas valley suggest that Zn is more mobile discussed here. scavenging of dissolved Zn to particulate

(that is, higher concentrations in We used a mass balance approach to Zn and sedimentation, depicted by RZn, porewater) under oxidizing than under examine various processes that control removes dissolved Zn from the water reducing conditions, most likely because the concentration of dissolved Zn in column of the lake. of formation of authigenic Zn sulfides Lake Coeur d’Alene (L.S. Balistrieri and Calculations were done using the under anoxic and sulfidic (reducing) P. F. Woods, unpublished data, 2002). computer program AQUASIM (Reichert, conditions. Lake Coeur d’Alene is treated as a 1994), which allows for simulations of single, completely mixed system (that is, aquatic systems. Discharge and dis- Modeling of coupled physical and a one-box model) (fig. 12). The follow- solved concentrations in the rivers for biogeochemical processes ing equation describes changes in dis- three water years (WY1999-2001; note Metal distributions and concen- solved Zn concentrations in the lake as a that water year 2000 runs from October trations in the environment are a result function of time (dCZn/dt): 1, 1999 to September 30, 2000) were of the interaction of many physical used (Brennan and others, 1999; transport processes and biogeochemical dCZn/dt = IZn + PZn - OZn - RZn (1) Brennan and others, 2000; Brennan and reactions. Understanding the dominant others, 2001) (fig. 13). Discharge during processes is critical for successful design where IZn accounts for all external inputs the 3 water years was variable; WY2000 of remediation activities and for long- of dissolved Zn to the lake, PZn accounts had 90 percent of the discharge of term management of ecosystems af- for all internal sources of dissolved Zn WY1999, whereas WY2001 had only 37 fected by human activities. Development to the lake, OZn accounts for fluxes of percent of the discharge of WY1999. of mathematical models that describe dissolved Zn out of the lake, and RZn Several assumptions were made in the metal cycling in such systems synthe- accounts for internal removal of dis- modeling: sizes our understanding of how the solved Zn from the water column. All 1. The inflow is equal to the outflow; system works by identifying and de- terms have units of mg/L per day. in other words, the volume of the scribing dominant processes and poten- Dissolved Zn enters the lake through lake does not change over time. tially allows for sensitivity tests to assess river inlets (Coeur d’Alene, St. Joe, and This assumption is a good approxi- how a system might respond to perturba- St. Maries Rivers) and leaves through a mation because the lake volume tions. Models can also assess our under- single outlet (Spokane River). These changes by less than 9 percent standing of system behavior by compar- processes are depicted by the IZn and OZn throughout the year (Woods and ing observations and predictions during terms in equation 1. Benthic fluxes of Beckwith, 1997).

10 Chapter 6 2. The volume of the lake is 2.8 x 1012 particulate Zn and subsequent just below the sediment-water interface L and the surface area of the lake- sedimentation is described by a in Lake Coeur d’Alene. The removal of 12 2 bottom sediment is 1.08 x 10 cm first-order rate coefficient (kscav), dissolved Zn from the water column is (Woods and Beckwith, 1997). and scavenging of dissolved Zn assumed to be by transformations to the 3. The exchange of dissolved Zn derived from the riverine input is particulate phase and settling of par- across the sediment-water interface the same as scavenging of dissolved ticles. No sediment trap data are avail- is by molecular diffusion; therefore, Zn derived from the benthic flux able to support this assumption. In a molecular diffusion coefficient is source. In other words, there is no addition, this removal is assumed to used to describe this process. This difference in the chemical behavior follow first-order kinetics. Limited assumption is supported by the or speciation of dissolved Zn from information from in situ microcosm work of Kuwabara and others the two sources. studies in lakes suggests that removal of (2000). The average value of the Model results depend on the formula- dissolved metals from the water column molecular diffusion coefficient of tion of the model (for example, the follows a first-order rate equation dissolved Zn at bottom water processes considered and their param- (Santschi and others, 1986; Anderson temperatures in Lake Coeur eterization) and the measured data that and others, 1987). d’Alene is 0.279 cm2 per day (Li are used. Assuming that all of the impor- Model results indicate that the relative and Gregory, 1974; Balistrieri, tant processes are incorporated into the importance of inputs of dissolved Zn to 1998). model, then the most reliable data are Lake Coeur d’Alene from rivers and 4. A range in the concentration of the physical characteristics of the system benthic flux varies through the year and dissolved Zn in the porewater just such as lake volume and surface area of from year to year (fig. 14A,B). The below the sediment-water interface sediment, the measured discharge and benthic flux contribution to dissolved Zn is used to calculate the input of dissolved Zn concentrations for the concentrations in the lake was larger dissolved Zn to the lake by benthic inflows and outflow, and the molecular than the river contributions during fluxes. Dissolved Zn concentrations diffusion coefficient for dissolved Zn. October of the first two water years of 1,070 (25th percentile), 1,210 The least well known data are the (WY99-WY00); otherwise, contribu- (median), and 1,650 (75th percen- porewater concentrations of dissolved tions from the inflowing river domi- tile) mg/L are used, based on 10 Zn, which determine the benthic flux, nated. During WY01, which was a low- published values (Balistrieri, 1998; and the parameterization of the removal discharge year, the supply of dissolved Kuwabara and others, 2000); these of dissolved Zn from the water column. Zn to the lake by benthic flux dominated concentrations do not vary season- Little information is available on spatial during the first 6 months (October ally in the model calculations. values and no information on seasonal through March), whereas the supply 5. Scavenging of dissolved Zn to values of dissolved Zn in the porewater from inflowing rivers was dominant

11 Chapter 6 during the subsequent 4 months (April solved Zn in Lake Coeur d’Alene. manganese dioxide. Because primary through July). Benthic flux was the What processes are responsible for productivity is seasonal, they observed primary input of dissolved Zn to the variations in the scavenging rate coeffi- corresponding seasonal variations in lake for the remainder of the water cient? Scavenging rate coefficients sedimentation of Zn. About 87 percent year. Low discharge results in less encompass the partitioning of metal of the dissolved Zn that enters Lake input of dissolved Zn to the lake from between dissolved and particulate Greifen is trapped within the lake, com- rivers and a higher proportion of input phases and the settling velocity or flux pared to 56 percent in Lake Coeur from benthic flux. In addition, the of particles (Sigg, 1985). Partitioning d’Alene. Sigg and others (1996) concluded relative importance of sources of between dissolved and particulate that biological mechanisms of uptake and dissolved Zn to the lake will change phases can be described by a distribu- binding are significant for the removal of as cleanup efforts within the mining tion coefficient and is a function of the dissolved Zn from the water column in district and lower Coeur d’Alene particle type, particle concentration, Lake Greifen. The largest scavenging rate River valley proceed. pH, ionic strength, and temperature of coefficients for dissolved Zn determined Values of the first-order scavenging the system. Scavenging rate coeffi- for Lake Coeur d’Alene occur during rate coefficient that are needed to fit cients are largest when dissolved metals periods when the maximum in biological the measured dissolved Zn concentra- have a strong affinity for particulate productivity would be expected. Further tions in the outflow (Spokane River) phases and the flux of particles in the work is needed to understand the mecha- range by a factor of 46, from 0.0005 system is large. Some insight into nisms and the role of biology in control- to 0.023 per day (fig. 14C). The variations of scavenging rate coeffi- ling transformations of dissolved to highest values of this coefficient were cients for Zn is given by work done in a particulate Zn in Lake Coeur d’Alene. determined in the third or fourth lake in Switzerland (Sigg and others, quarter of the water year and slightly 1996). Although they examined a precede the minimum in dissolved Zn relatively pristine lake relative to Lake What is known about impacts of concentration. For comparison, values Coeur d’Alene (Lake Greifen, Switzer- Pb and Zn to biota, and can spe- of scavenging coefficients for dis- land, where Zn concentrations are 0.65- cific processes that influence solved Zn from whole-lake experi- 2.6 mg/L), the chemical behavior of Zn bioavailability be identified? ments and microcosm studies using should be similar in the two lakes. radiotracers range from 0.03 to 0.06 Using sediment trap data, they found The health impacts of mining in the per day (Santschi and others, 1986; that the sedimentation of Zn (or trans- Coeur d’Alene River basin on humans Anderson and others, 1987). These formations from dissolved to particu- and other biota are highly dependent on values are at the high end of values late phases and subsequent settling) was the phase of the metal. The critical determined by the modeling of dis- related to sedimentation of algae and phase for Pb is the particulate form,

12 Chapter 6 whereas for Zn it is the dissolved form. between the ages of 9 months and 9 bioavailability of Pb are linked to the Some of the highest levels of Pb in years have higher than recommended speciation of Pb in the solid phase, the blood of children in the United levels of Pb in their blood which depends on the redox conditions States were measured in the Coeur (TerraGraphics Environmental Engi- of the depositional environment. The d’Alene mining district around the neering and others, 2001). Warnings impact of Zn on biota depends on the Bunker Hill smelter and the eventual of potential human exposure to high mobilization of this element from Superfund site (fig. 15). The primary concentrations of metals in sediment metal-enriched tailings and subsequent pathway of Pb into human blood is and fish and related health impacts transport and reaction between ground ingestion of Pb-enriched particles are posted at recreational beaches water and surface water. that are soluble in the stomach. along the lower Coeur d’Alene River, Solubility of the particles depends on Coeur d’Alene Lake, and Spokane What are the major gaps in the speciation of Pb in the solid River. knowledge remaining from this phase (Ruby and others, 1992; Gas- High concentrations of Pb in surface research? ser and others, 1996). In 1974, one sediments within the basin and dis- year after a fire destroyed the air solved Zn in rivers also have proved 1. Quantitative and predictive dynamic emission controls on the Bunker Hill harmful to terrestrial and aquatic models are needed that provide an smelter stack, 98 percent of children wildlife. The metal-enriched marshes understanding of the underlying 1 to 9 years old in the area around within the lower Coeur d’Alene River mechanisms controlling the trans- the smelter had blood Pb levels >40 valley are prime resting and feeding port, reaction, and fate of Pb, Zn, mg/dL (U.S. Environmental Protec- areas for migratory birds, and deaths and other elements in the Coeur tion Agency, 1994). Emergency related to Pb poisoning during feeding d’Alene River basin and other large- response actions (for example, edu- in this area have been reported for scale systems affected by mining cation about hygiene, smelter clo- waterfowl (Beyer and others, 1998). activities. Sensitivity tests with such sure, and yard remediation) resulted High dissolved Zn concentrations in models would provide information in lower blood Pb levels, most of the South Fork of the Coeur d’Alene on the response of large-scale which are now below the maximum River and its tributaries have prevented systems to natural and human- recommended level of 10 mg/dL. the re-establishment of a viable fish- induced perturbations, would iden- However, recent studies of blood Pb ery, despite attempts to improve fish tify data gaps, and could be used to levels in children living outside of habitat in streams (T. Maret, unpub- direct remediation and management the Superfund site, but within the lished data, 1999). of the systems. Coeur d’Alene River basin, indicate Factors that affect the 2. The interrelationships among that about 15 percent of children geoavailability and, ultimately, processes and factors responsible

13 Chapter 6 for chemical speciation, mobiliza- solved metals in Coeur d’Alene polymetallic vein deposits; a case tion, and transport of Pb, Zn, and Lake, Idaho: U.S. Geological Sur- study in the Coeur d’Alene mining other elements in the environment vey Open-File Report 98-793, 40 p. district and comparisons with (geoavailability) and the uptake of Balistrieri, L.S., Bookstrom, A.A., drainage from mineralized deposits Pb, Zn, and other elements by biota Box, S.E., and Ikramuddin, M., in the Colorado Mineral Belt and (bioavailability) need to be further 1998, Drainage from adits and tail- Humboldt Basin, Nevada, in Seal, defined. ings piles in the Coeur d’Alene R.R., II., and Foley, N.K., eds., 3. The role of microorganisms in mining district, Idaho; sampling, Progress on geoenvironmental determining chemical speciation, analytical methods, and results: models of mineral deposits, U.S. geoavailability, and bioavailability U.S. Geological Survey Open-File Geological Survey Open-File Re- for Pb, Zn, and other elements Report 98-127, 19 p. port 02-195, p. 143-160. needs to be further evaluated. Balistrieri, L.S., Box, S.E., Barton, G.J., 2002, Dissolved cad- Bookstrom, A.A., and Ikramuddin, mium, zinc, and lead loads from ACKNOWLEDGMENTS M., 1999, Assessing the influence ground-water seepage into the Robert Seal, Michael Zientek, Tom of reacting pyrite and carbonate South Fork Coeur d’Alene River Frost, and Peter Vikre of the U.S. Geo- minerals on the geochemistry of system, Northern Idaho, 1999: U.S. logical Survey provided comments on an drainage in the Coeur d’Alene min- Geological Survey Water-Re- earlier draft of this paper. Peter Stauffer’s ing district: Environmental Science sources Investigations Report 01- editing of the manuscript helped im- and Technology, v. 33, no. 19, p. 4274, 139 p. prove its clarity. 3347-3353. Bennett, E.H., and Venkatakrishnan, Balistrieri, L.S., Box, S.E., Ikramuddin, R., 1982, A palinspastic recon- REFERENCES M., Horowitz, A.J., and Elrick, struction of the Coeur d’Alene Anderson, R.F., Santschi, P.H., Nyffeler, K.A., 2000, A study of porewater in mining district based on ore depos- U.P., and Schiff, S.L., 1987, Validat- water saturated sediments of levee its and structural data: Economic ing the use of radiotracers as analogs banks and marshes in the lower Geology, v. 77, p. 1851-1866. of stable metal behaviour in enclosed Coeur d’Alene River valley, Idaho; Bennett, E.H., Siems, P.L., and aquatic ecosystem experiments: Ca- sampling, analytical methods, and Constantopoulos, J.T., 1989, The ge- nadian Journal of Fisheries and results: U.S. Geological Survey ology and history of the Coeur Aquatic Science, v. 44, Supplement Open-File Report 00-126, 62 p. d’Alene mining district, Idaho, in 1, p. 251-259. Balistrieri, L.S., Box, S.E., and Chamberlain, V.E., Breckenridge, Balistrieri, L.S., 1998, Preliminary es- Bookstrom, A.A., 2002, A R.M., and Bonnichsen, B., eds., timates of benthic fluxes of dis- geoenvironmental model for Guidebook to the Geology of

14 Chapter 6 Northern and Western Idaho and vey Open-File Report 02-126, 76 p. port ID-99-2, 440 p. surrounding area, Idaho Geological Box, S.E., Bookstrom, A.A., Brennan, T.S., Campbell, A.B., Survey Bulletin 28, p. 137-156. Ikramuddin, M., and Lindsay, J., Lehmann, A.K., and O’Dell, I., Beyer, W.N., Audet, D.J., Morton, A., 2001, Geochemical analyses of 2001, Water resources data, Idaho, Campbell, J.K., and LeCaptain, L., soils and sediments, Coeur d’Alene water year 2000, Volume 2. Upper 1998, Lead exposure of waterfowl drainage basin, Idaho; sampling, Columbia River Basin and Snake ingesting Coeur d’Alene River ba- analytical methods, and results: U. River Basin below King Hill: U. S. sin sediment: Journal of Environ- S. Geological Survey Open-File Geological Survey Water-Data Re- mental Quality, v. 27, no. 6, p. Report 01-139, 206 p. port ID-00-2, 402 p. 1533-1538. Box, S.E., Bookstrom, A.A., and Cravotta, C.A., III, Brady, K.B.C., Bookstrom, A.A., Box, S.E., Jackson, Ikramuddin, M., in press, Metal- Smith, M.W., and Beam, R.L., B.L., Brandt, T.R., Derkey, P.D., and enriched sediments mobilized by 1990, Effectiveness of alkaline ad- Munts, S.R., 1999, Digital map of the floods of 1995, 1996, and 1997 dition at surface mines in prevent- surficial geology, wetlands, and in the Coeur d’Alene-Spokane ing or abating acid mine drainage deepwater habitats, Coeur d’Alene River drainage, Idaho and Wash- Part 1. geochemical considerations, River valley, Idaho: U.S. Geological ington: U.S. Geological Survey in Skousen, J.G., Sencindiver, J., Survey Open-File Report 99-548, Open-File Report. and Samuel, D.E., eds., Proceed- 121 p. Brennan, T.S., Lehmann, A.K., O’Dell, ings of the 1990 Mining and Recla- Bookstrom, A.A., Box, S.E., Campbell, I., and Tungate, A.M., 1999, Water mation Conference and Exhibition, J.K., Foster, K.I., and Jackson, B.L., resources data, Idaho, water year p. 221-226. 2001, Lead-rich sediments, Coeur 1998 Volume 2. Upper Columbia Criss, R.E., and Fleck, R.J., 1990, d’Alene River valley, Idaho; area, River Basin and Snake River Basin Oxygen isotopic map of the giant volume, tonnage, and lead content: below King Hill: U.S. Geological metamorphic-hydrothermal system U.S. Geological Survey Open-File Survey Water-Data Report ID-98- around the northern Idaho Report 01-140, 57 p. 2, 386 p. batholith, U.S.A.: Applied Box, S.E., and Wallis, J.C., 2002, Brennan, T.S., Campbell, A.M., Geochemistry, v. 5, p. 641-655. Surficial geology along the Spo- Lehmann, A.K., and O’Dell, I., Dzombak, D.A., and Morel, F.M.M., kane River, Washington, and its re- 2000, Water resources data, Idaho, 1990, Surface complexation mod- lationship to the metal content of water year 1999 Volume 2. Upper eling hydrous ferric oxide: New sediments (Idaho-Washington Columbia River Basin and Snake York, John Wiley & Sons, 393 p. stateline to Latah Creek River Basin below King Hill: U.S. Froelich, P.N., Klinkhammer, G.P., confluence): U. S. Geological Sur- Geological Survey Water-Data Re- Bender, M.L., Luedtke, N.A.,

15 Chapter 6 Heath, G.R., Cullen, D., Dauphin, Gott, G.B., and Cathrall, J.B., 1980, and related activities on the sedi- P., Hammond, D.E., Hartman, B., Geochemical exploration studies in ment trace element geochemistry and Maynard, V., 1979, Early oxi- the Coeur d’Alene district, Idaho of Lake Coeur d’Alene, Idaho, dation of organic matter in pelagic and Montana: U. S. Geological USA. Part I. surface sediments: sediments of the eastern equatorial Survey Professional Paper 1116, 63 Hydrological Processes, v. 7, p. Atlantic; suboxic diagenesis: p. 403-423. Geochimica et Cosmochimica Hobbs, S.W., Griggs, A.B., Wallace, Horowitz, A.J., Elrick, K.A., Acta, v. 43, p. 1075-1090. R.E., and Campbell, A.B., 1965, Robbins, J.A., and Cook, R.B., Fryklund, V.C., Jr., 1964, Ore deposits Geology of the Coeur d’Alene dis- 1995, Effect of mining and re- of the Coeur d’Alene district, trict, Shoshone County, Idaho: U.S. lated activities on sediment trace Shoshone County, Idaho: U.S. Geo- Geological Survey Professional Pa- element geochemistry of Lake logical Survey Professional Paper per 478, 139 p. Coeur D’Alene, Idaho, USA. Part 445, 103 p. Hooper, R.L., and Mahoney, J.B., II. subsurface sediments: Hydro- Gasser, U.G., Walker, W.J., Dahlgren, 2000, Constraining contaminant logical Processes, v. 9, p. 35-54. R.A., Borch, R.S., and Burau, transport; lead and zinc speciation Hudson-Edwards, K.A., Macklin, R.G., 1996, Lead release from in fluvial subenvironments, lower M.G., Curtis, C.D., and Vaughan, smelter and mine waste impacted Coeur d’Alene River valley, Idaho: D.J., 1996, Processes of forma- materials under simulated gastric Geological Society of America Ab- tion and distributions of Pb-, Zn- conditions and relation to specia- stracts with Programs, v. 32, no. 7, , Cd-, and Cu-bearing minerals in tion: Environmental Science and p. 125. the Tyne Basin, Northeast En- Technology, v. 30, no. 3, p. 761- Hooper, R.L., and Mahoney, J.B., gland; implications for metal- 769. 2001, Metal transport, heavy metal contaminated river systems: En- Gitlin, E., 1986, Wall rock geochemis- speciation and microbial fixation vironmental Science and try of the Lucky Friday Mine, through fluvial subenvironments, Technology, v. 30, no. 1, p. 72- Shoshone County, Idaho: Seattle, lower Coeur d’Alene River valley, 80. University of Washington, Ph.D. Idaho: Eos, Transactions, American Kuwabara, J.S., Berelson, W.M., dissertation, 224 p. Geophysical Union, v. 82, no. 47, Balistrieri, L.S., Woods, P.F., Golder Associates, 1998, Hydrological supplement (Abstracts of AGU Topping, B.R., Steding, D.J., and investigation, Interstate Mill, 2001 fall meeting), p. F199- Krabbenhof, D.P., 2000, Benthic Wallace, Idaho: Report to Silver F200. flux of metals and nutrients into Valley Natural Resource Trustees, Horowitz, A.J., Elrick, K.A., and the water column of Lake Coeur 10 p. Cook, R., 1993, Effect of mining d’Alene, Idaho; report of an Au-

16 Chapter 6 gust, 1999, pilot study: U.S. Geo- Shoshone County, Idaho; Pre- geochemical calculations: U.S. logical Survey Water-Resources liminary Estimates: U.S. Geo- Geological Survey Water-Re- Investigations Report 00-4132, logical Survey Open-File Report sources Investigations Report 99- 74 p. 98-595, 14 p. 4259, 312 p. Leach, D.L., Landis, G.P., and McCulley Frick and Gilman, 1994, Paulson, A.J., and Balistrieri, L.S., Hofstra, A.H., 1988, Metamorphic Bunker Hill Superfund site, 1993 1999, Modeling removal of Cd, Cu, origin of Coeur d’Alene base- and annual groundwater monitoring Pb, and Zn in acidic groundwater precious-metal veins in the Belt report: Basic data report, 41 p. during neutralization by ambient basin, Idaho and Montana: Geol- McCulley Frick and Gilman, 2000, surface waters and groundwaters: ogy, v. 16, p. 122-125. 1999 annual water-quality report, Environmental Science and Tech- Leach, D.L., Hofstra, A.H., Church, Woodland Park, Idaho: Consult- nology, v. 33, no. 21, p. 3850-3856. S.E., Snee, L.W., Vaughn, R.B., ant report prepared for Hecla Plumlee, G.S., Smith, K.S., Ficklin, and Zartman, R.E., 1998, Evi- Mining Co. and Re- W.H., Briggs, P.H., and McHugh, dence of Proterozoic and Late source Trustees, 24 p. J.B., 1993, Empirical studies of di- Cretaceous-early Tertiary ore- Mink, L.L., Williams, R.E., and verse mine drainages in Colorado; forming events in the Coeur Wallace, A.T., 1971, Effect of in- implications for the prediction of d’Alene district, Idaho and Mon- dustrial and domestic effluents mine-drainage chemistry, in 1993 tana: Economic Geology, v. 93, p. on the water quality of the Coeur Mined Land Reclamation Sympo- 347-359. d’Alene River basin: Idaho Bu- sium, Billings, Mont.,1993, Pro- Li, Y.-H., and Gregory, S., 1974, reau of Mines and Geology Pam- ceedings: Billings, Mont., Montana Diffusion of ions in sea water phlet 149, 30 p. State University Reclamation Re- and in deep-sea sediments: Nash, J.T., 2000, Hydrogeochemical search Unit, p. 176-186. Geochimica et Cosmochimica data for historic mining areas, Plumlee, G.S., Smith, K.S., Montour, Acta, v. 38, p. 703-714. Humboldt watershed and adja- M.R., Ficklin, W.H., and Mosier, Long, K.R., 1998a, Grade and ton- cent areas, northern Nevada: U.S. E.L., 1999, Geologic controls on nage models for Coeur d’Alene- Geological Survey Open-File Re- the composition of natural waters type polymetallic veins: U.S. Geo- port 00-459, 13 p. and mine waters draining diverse logical Survey Open-File Report Parkhurst, D.L., and Appelo, C.A.J., mineral-deposit types, in Filipek, 98-583, 28 p. 1999, User’s Guide to PHREEQC L.H., and Plumlee, G.S., eds., The Long, K.R., 1998b, Production and (Version 2)-a computer program environmental geochemistry of min- disposal of mill tailings in the for speciation, batch-reaction, one- eral deposits, Part B. case studies Coeur d’Alene mining region, dimensional transport, and inverse and research topics, reviews in eco-

17 Chapter 6 nomic geology, Society of Economic try, v. 1, p. 313-328. mental Science and Technology, v. 36, Geologists, p. 373-432. Ter raGraphics Environmental Engi- no. 3, p. 484-492. Reichert, P., 1994, AQUASIM-a tool for neering, 1998, Preliminary engi- U.S. Environmental Protection Agency, simulation and data analysis of neering design report, Success 1994, Cleanup of the Bunker Hill aquatic systems: Water Science mine site passive treatment project: Superfund site; an overview: U.S. Technology, v. 30, no. 2, p. 21-30. Prepared for Idaho Department of Environmental Protection Agency Ruby, M.V., Davis, A., Kempton, J.H., Health and Welfare, Division of Report EPA 910-R-94-009, 10 p. Drexler, J.W., and Bergstrom, P.D., Environmental Quality, 25 p. U.S. Environmental Protection Agency, 1992, Lead bioavailability; dissolu- TerraGraphics Environmental Engineer- 1999, National Recommended Water tion kinetics under simulated gastric ing, URS Greiner, and CH2M Quality Criteria-correction: U.S. En- conditions: Environmental Science HILL, 2001, Human health risk as- vironmental Protection Agency Re- and Technology, v. 26, no. 6, p. 1242- sessment for the Coeur d’Alene ba- port EPA 822-Z-99-001, 25 p. 1248. sin extending from Harrison to U.S. General Services Administration, Santschi, P.H., Nyffeler, U.P., Anderson, Mullan on the Coeur d’Alene River 1999, Code of Federal Regulations, R.F., Schiff, S.L., and O’Hara, P., and tributaries, remedial investiga- Title 40, Part 131.36. 1986, Response of radioactive trace tion/feasibility study: Prepared for U.S. Geological Survey, 1973, Water Re- metals to acid-base titrations in con- Idaho Department of Health and sources data for Idaho in 1972, Part trolled experimental ecosystems; Welfare, Division of Health, Idaho 2, water quality records: U.S. Geo- evaluation of transport parameters for Department of Environmental Qual- logical Survey Water-Data Report applications to whole-lake radiotracer ity, and U.S. EPA Region 10, 392 p. ID-72-2, 239 p. experiments: Canadian Journal of (+ appendices). Woods, P., F., and Beckwith, M.A., 1997, Fisheries and Aquatic Science, v. 43, Tessier, A., Campbell, P.G.C., and Nutrient and trace-element enrich- p. 60-77. Bisson, M., 1979, Sequential extrac- ment of Coeur d’Alene Lake, Idaho: Sigg, L., 1985, Metal transfer mecha- tion procedure for the speciation of U.S. Geological Survey Water-Sup- nisms in lakes; the role of settling particulate trace metals: Analytical ply Paper 2485, 93 p. particles, in Stumm, W., ed., Chemi- Chemistry, v. 51, p. 844-851. Zartman, R.E., and Stacey, J.S., 1971, cal Processes in Lakes: New York, Tonkin, J.W., Balistrieri, L.S., and Lead-isotopes and mineralization John Wiley & Sons, p. 283-310. Murray, J.W., 2002, Modeling metal ages in Belt Supergroup rocks, Sigg, L., Kistler, D., and Ulrich, M.M., removal onto natural particles formed northwestern Montana and northern 1996, Seasonal variations of zinc in a during mixing of acid rock drainage Idaho: Economic Geology, v. 66, p. eutrophic lake: Aquatic Geochemis- with ambient surface water: Environ- 849-860.

18 Chapter 6 116°W Spokane River N AREA OF MAP 0 10 MILES Coeur d'Alene Washington Montana 0 10 KILOMETERS North Fork

Oregon Idaho Wyoming Coeur d'Alene River Coeur d'Alene Kellogg Coeur Cataldo Smelterville Rose mining district d'Alene Lake Lake SouthFork Burke Kingston ° 47.5 N Elizabeth Osburn Mullan Park Wallace Harrison Bunker Hill Superfund site

St. Joe River

Figure 1.—Index map of the Coeur d’Alene River basin. The dark squares are towns.

19 Chapter 6 Figure 2.—Annual production in metric tons and percent metal content of tailings produced at the Morning Mill from 1895 to 1958. Unpublished data from K. Long, U.S. Geological Survey.

20 Chapter 6 SEDIMENT LAYERS AND ASSOCIATED LEAD CONTENTS a Medium-grained sand: 3-4,000 ppm Pb b Interbedded fine sand and silt: 4-10,000 ppm Pb c Silt and very fine sand: 15-36,000 ppm Pb d Pre-mining sand and silt: < 30 ppm Pb

Channel of Coeur d'Alene River East West b Summer water level Winter water level

a b d c

5 meters Core location 10 meters

Figure 3.—Stylized cross section showing the Pb content of sediments in cores from the floodplain and channel of the Coeur d’Alene River near Killarney (between Rose Lake and Harrison). Data from Box and others (2001).

21 Chapter 6 2,500

2,500 4,000 2,000 5,000

1,500

1,000

500

0 051015

Figure 4.—Pb concentrations and Zn/Pb ratios in the <175 mm fraction of bed sediment in the Spokane River compared to results of a mixing model involving three different sources of sediment.

22 Chapter 6 Figure 5.—Dissolved Zn concentrations and pH values in river water, ground water, porewater, tailings seeps, and adit drainage in the Coeur d’Alene River basin. Also shown are water quality criteria for pH and Criterion Continuous Concentration (CCC) for Zn, which is the highest concentration of total recoverable Zn to which aquatic life can be exposed for an extended period of time (4 days) without deleterious effects in Idaho water. See text and table 1 for data sources.

23 Chapter 6 Figure 6.—Scanning electron microscope (SEM) im- age and X-ray intensity maps of a single detrital sili- cate grain coated with Pb/Fe/Mn oxide from the surficial floodplain soil in Smelterville Flats. A, SEM image. B, X-ray intensity map of Pb. C, X-ray inten- Figure 7.—Calibration of a sequential extraction scheme based on the method of Tessier sity map of Mn. Images suggest that Pb is primarily and others (1979) by mineralogical analyses after each extraction step of particles associated with Mn oxide. collected in the Coeur d’Alene River basin.

24 Chapter 6 Figure 8.—Sequential extraction results for samples collected in the river channel, on top of levee banks, and in wetlands in the lower Coeur d’Alene River valley.

25 Chapter 6 Figure 9.—Comparisons of the Pb and Zn contents of suspended sediments during the spring floods in 1996 and 1997 in the Coeur d’Alene River basin. The 1996 flood was larger than the 1997 flood.

26 Chapter 6 Figure 10.—Comparisons between the predicted ratio of reacting pyrite-to-calcite minerals (considering pre- cipitation of ferrihydrite and calculated as pyritic SO4/(Ca + Mg) in drainage) and pH in drainage from polymetallic veins in the Coeur d’Alene mining district, the Colorado Mineral Belt, and Humboldt River Basin, Nevada. See text for data sources.

27 Chapter 6 Figure 11.—Comparison of measured and predicted values of pH, and of Pb adsorbed on Fe precipitates formed, during the mixing of Coeur d’Alene River water and porewater from metal-enriched sediments near the river’s edge in Cataldo, Smelterville Flats, Killarney, and Rose Lake. Errors bars on measured particulate Pb data are propagated error in calculating particulate concentrations from initial and final dissolved Pb concentrations in the mixing experiments. Model predictions were made using PHREECQC (Parkhurst and Appelo, 1999) and a sorption database for ferric oxide (Dzombak and Morel, 1990).

28 Chapter 6 Figure 12.—A one-box model depicting the processes considered in modeling dissolved Zn concentrations in Lake Coeur d’Alene.

29 Chapter 6 Figure 13.—Discharge data and dissolved Zn concentrations for rivers entering and leaving Lake Coeur d’Alene during water years (WY) 1999-2001. Concentrations of dissolved Zn in the St. Joe and St. Maries Rivers are not shown and typically were below published detection limits of <1 or <20 mg/L (note: detection limits varied over time); when detected, concentrations in these rivers were <4 mg/L.

30 Chapter 6 /L /L /L /L /L /L P k S

Figure 14.—Results of box model calculations for dissolved Zn in Lake Coeur d’Alene. A, Comparisons of measured and modeled dissolved Zn concentrations as a function of water years (WY) 1999 through 2001, taking into account variable discharge and dissolved Zn concentra- tions in the inflow and outflow and benthic fluxes. Concentrations of dissolved Zn as a function of time in the St. Joe and St. Maries Rivers were assumed to be constant at a value of 1.5 mg/L for the model runs. B, Relative contributions of dissolved Zn to the lake from inflowing rivers and from benthic flux. C, Model values for the scavenging rate coefficient (kscav) that fit observed dissolved Zn concentrations in the outflow. 31 Chapter 6 Figure 15.—Concentrations of Pb in the blood of people living within the Bunker Hill Superfund site as a function of time. Data from U.S. Environmental Protection (EPA) web page (www.epa.gov), search on Bunker Hill Superfund. The U.S. Environmental Protection Agency (EPA) and Centers for Disease Control and Prevention (CDC) determined that concentrations of Pb in blood at or above 10 mg/dL present risks to the health of children.

32 Chapter 6 Table 1.—Range and median values of pH and dissolved Zn concentrations in waters in the Coeur d’Alene (CdA) River Basin.

[Data sources: Mink and others, 1971; U.S. Geological Survey, 1973; McCulley Frick and Gillman, 1994; Balistrieri and others, 1998; Golder Associates, 1998; Terragraphics Environmental Engineering, 1998; S.E. Box, unpublished data, 1999; Balistrieri and others, 2000; McCulley Frick and Gilman, 2000.] Water type Number pH Dissolved Zn (mg/L) of samples Range Median Range Median

Ground waters 246. 3.5-7.6 5.61 0.18-759 38.

Tailings seeps 26. 3.8-8.2 5.93 0.09-498 66.

Porewaters 30. 6.2-7.1 6.57 0.005-70 10.

Adits: Kellogg Tunnel 1. 2.7 615 All other adits 61. 5.5-8.3 7.34 0.001-58 5.8

CdA River 117. 4.4-9.1 7.05 0.05-21 3.4

33 Chapter 6 Table 2.—Summary of mineralogical and leach studies of particles in the Coeur d’Alene mining district. Location Smelterville Flats Smelterville Flats Osburn Flats Osburn Flats Upper Woodland Park Upper Woodland Park Environment floodplain channel floodplain channel floodplain channel Sample number 201 202 203 204 205 206 Total Pb (ppm) 18,500. 5,760. 34,600. 4,665. 123,800. 5,670. Total Fe (%) 16. 5.3 17. 7.9 14. 7.3 Total Mn (ppm) 11,720. 3,765. 17,530. 5,160. 2,485. 3,275. SEM/EDS and XRD1 Bulk - metallic minerals Major dispersed Pb, Mn-Pb ox Pb-Mn ox Cer, Pyr Ang Pb-Fe ox, dispersed Pb Minor Mn-Pb ox, Pb-Mn-Fe ox Zn-Pb-Fe ox Pb-Fe ox, Mn-Pb ox Fe-Pb ox, dispersed Pb Trace Sph, Cer, Pyr Sid, Cer, Sph Sid, Pyr, Sph, Cer Sid, Cer, Sph Sph Sid, Cer, Sph, Ang Sequential extraction2

% Pb in MgCl2 leach 4.6 13. 5.5 3.3 39. 9.8 % Pb in NaOH leach 1.7 13. 1.6 11. 17. 16. % Pb in EDTA leach 18. 35. 16. 35. 32. 35.

% Pb in HNO 3 leach 78. 38. 75. 50. 10. 40. Gastric leach3 % Pb 44. 73. 39. 61. 16. 76. % Fe .7 4.7 .7 2.3 .8 3.5 % Mn 4.5 26. 5.5 17. 1.2 30.

1 Mineral abbreviations: ox = oxide, Sph = sphalerite (ZnS), Cer = cerrusite (PbCO3), Pyr = pyrite (FeS2), Sid = siderite (FeCO3), Ang = anglesite (PbSO4) 2 Sequential extraction leaches (Gasser and others, 1996) 1 M MgCl2: easily soluble minerals such as Pb sulfate (Ang) 0.5 M NaOH: organically complexed Pb and Pb carbonates (Cer) 0.05 M Na2EDTA: specifically adsorbed Pb, Pb oxides, Pb phosphates, and some amorphous Fe/Mn oxides 4 M HNO : Pb sulfide (galena) and remainder of Fe/Mn oxides 3 3 Gastric leach: 0.1M HCl/0.1M NaCl for 1 hr at 37°C (Gasser and others, 1996)

34 Chapter 6