U.S. Geological Survey Toxic Substances Hydrology Program--Proceedings of the Technical Meeting, Colorado Springs, Colorado, September 20-24, 1993, Water-Resources Investigations Report 94-4015

Hydrologic and Geochemical Factors affecting Metal Contaminant Transport in Pinal Creek Basin near Globe, Arizona

Contents

• ABSTRACT
• INTRODUCTION
• GEOHYDROLOGY
• NATURE AND EXTENT OF CONTAMINANTION
• RESEARCH IN PROGRESS
  • Ground-Water Geochemistry and Transport
  • Ground-Water Hydrology
  • Metal Transport in Shallow Ground Water and Streamflow
• SUMMARY
• REFERENCES CITED

ABSTRACT

Acidic-mine drainage from copper mining in the Pinal Creek basin has contaminated the regional aquifer. Chemical reactions neutralize the contaminant plume as it moves downgradient. Neutralized contamination subsequently surfaces in a perennial reach of Pinal Creek and equilibrates with the atmosphere, causing pH to rise and manganese minerals to precipitate on the streambed. Ongoing research at the site is focused on ground-water hydrology and geochemistry, metal transport, and interactions between shallow ground water and streamflow.

Field and laboratory experiments will attempt to identify the reduction(s) that balance the oxidation of iron in the aquifer. Reactive-transport simulations of plume evolution and remediation will identify the important processes and reactions that produced the observed conditions in the plume. Measurements of gravity change and ground-water level change at project wells may provide refined estimates of aquifer specific yield for use in reactive-transport simulations. Estimates of ground-water ages by chlorofluorocarbon age-dating techniques may improve estimates of traveltimes and reaction rates of contaminants in ground water.

Mechanisms of trace-metal scavenging by manganese carbonates, manganese oxides, and iron oxides in the perennial stream are being investigated, and partitioning coefficients are being estimated to constrain hydrogeochemical modeling. A coupled aquifer-stream mass-balance approach has been used to constrain effective rates of solid-aqueous partitioning of manganese, iron, nickel, copper, and cobalt. Tracer experiments are being undertaken to distinguish reactions occurring in near-stream ground water from reactions in streamflow.

INTRODUCTION

The Pinal Creek basin, in central Arizona (fig. 1), has been an area of large-scale copper mining since the late 1880's. Acidic drainage from mining and related activities has contaminated the regional alluvial aquifer. The U.S. Geological Survey began a study of the contamination with the following objectives: (1) Define the nature and extent of ground-water and surface-water contamination, (2) identify the reactions and processes occurring in the contaminant plume, and (3) develop and test the ability of geochemical and transport models to simulate observed conditions. The purpose of this paper is to provide an overview of the site and a summary of past and ongoing research at Pinal Creek.

(99k)

Figure 1. Area of study

GEOHYDROLOGY

The drainage area of Pinal Creek above Inspiration Dam is 516 km². About 170 km² of this area is underlain by a basin-fill conglomerate and a permeable stream alluvium, which together compose the regional aquifer. The aquifer is bounded laterally and at depth in most areas by relatively impermeable rocks. The aquifer thins and becomes narrower from well site 500 (fig. 2) north to Inspiration Dam. This thinning and narrowing decreases the volume of water that can move and be stored in the subsurface, and sustains perennial flow into Pinal Creek. Streamflow in Pinal Creek is perennial in a 5-to 6 kilometer reach above Inspiration Dam.

Basin fill ranges from unsorted and unconsolidated boulders to well-stratified deposits of firmly cemented sand, silt, and gravel (Peterson, 1962, p. 41). Carbonate content of the basin fill is about 1.5 percent (Eychaner, 1989, p. 570). Horizontal hydraulic conductivity of basin fill is estimated to range from 0.1 to 0.2 m/day (C.G. Taylor, Magma Copper Corporation., written commun., 1987). Stream alluvium consists mainly of sand and gravel but includes stratified deposits of clay-sized material as much as 12 m thick locally. Horizontal hydraulic conductivity of alluvium between well sites 400 and 500 was estimated to be 260 m/d. Alluvium at well site 600 (fig. 2), through which neutralized, contaminated water has passed, contained from 0.3- to 0.7- percent carbonate material.

(33k)

Figure 2. Generalized hydrogeologic section of the aquifer. Line of section approximates the principal ground-water flow line from Webster Lake to Inspiration Dam along the channels of Webster Gulch, Bloody Tanks Wash, Miami Wash, and Pinal Creek.

Ground-water levels in stream alluvium and adjacent basin fill respond rapidly to variations in recharge and pumpage. Water levels declined steadily from 1986 to 1990 in response to less-than-average rainfall in the basin and increased pumping between Miami Wash and Pinal Creek for the purposes of remediation. The amount of ground-water decline generally decreased in a downstream direction. Ground-water levels rose as much as 16 m (fig. 3) adjacent to Miami Wash by February 1993 because of wetter than normal winters in 1991 and 1993.

(33k)

Figure 3. Ground-water levels and concentrations of iron and sulfate adjacent to Miami Wash.

NATURE AND EXTENT OF CONTAMINANTION

Leakage of mine wastewater and process water from unlined impoundments into stream alluvium have contributed to a 15-kilometer-long acidic contaminant plume within the alluvium along Miami Wash and Lower Pinal Creek (fig. 4). Webster Lake, the single largest potential source of contamination, was used to store waste and process water from 1941 until 1988, when it was drained by the order of the U.S. Environmental Protection Agency. Contaminated ground water removed by remedial pumping is either spread on mine tailings to evaporate or is used at the mines.

(33)

Figure 4. Distribution of pH in the aquifer in 1992. Line of section approximates the principal ground-water flow line from Webster Lake to Inspiration Dam along the channels of Webster Gulch, Bloody Tanks Wash, Miami Wash, and Pinal Creek.

Water from well 51, which is completed in the acidic core of the plume, contained more than 500 mg/L iron (Fe), 32 mg/L copper (Cu), and 2,700 mg/L sulfate (SO₄) at a pH of 3.7 in 1992 (table 1). Between site 450 and the perennial reach of Pinal Creek, neutralization reactions caused pH of ground water to increase from between 4 and 5 to about 6 in 1992. In February 1993, neutralized ground water discharged to a perennial reach that began less than 1 km above well site 500. As the neutralized ground water discharges into the stream, gases dissolved in water reach equilibrium with the atmosphere. In the process, carbon dioxide (CO₂) degases, and pH rises to about 8 over a distance of about 5 km. Neutralized water from Pinal Creek at Inspiration Dam (Streamflow-gaging station 09498400) contained 560 mg/L dissolved calcium (Ca), 2,800 mg/L dissolved SO₄, and 39 mg/L dissolved manganese (Mn) in May 1992. Dissolved Fe and Cu concentrations were 0.01 and 0.006 mg/L (micrograms per liter), respectively. From 1984 to 1992, concentrations of Fe in water from well 51 decreased from 3,100 mg/L to 500 mg/L, and concentrations of SO₄ decreased from 10,000 mg/L to 2,700 mg/L (fig. 3). These decreases were caused by (1) removal of a significant source of contamination, (2) recharge from winter runoff in 1991-92, and (3) chemical reactions in the subsurface.

RESEARCH IN PROGRESS

Ongoing research falls within three general categories: ground-water geochemistry and transport, ground-water hydrology, and hydrochemical interactions between ground water and surface water. These broad categories provide the framework in which to discuss present research at the site.

Ground-Water Geochemistry and Transport

Although the most important reactions in the aquifer are well understood, the redox reactions responsible for the observed decreases in dissolved Fe have not been adequately characterized. The results of geochemical modeling (Glynn and others, 1991) suggest that the dissolution of Mn oxide solids could be the principal reduction that balances Fe oxidation. Reactions involving Mn and Fe indicate that 2 mol of Fe should precipitate for each mole of Mn that dissolves, even though column experiments by Stollenwerk gave the ratio of 7 to 1 (Eychaner, 1991a). This ratio suggests that (1) Mn-oxide solids are reduced incongruently to form other Mn minerals, (2) other minerals or aqueous species might also be reduced, and (3) the atmosphere may be a significant source of O₂ to the plume. The incongruent dissolution of Mn-oxide solids and potential reduction of other minerals or aqueous species to balance the redox budget will be examined by a combination of laboratory work and geochemical modeling.

Table 1. Chemical analyses of contaminated water in Pinal Creek basin
[Concentrations in milligrams per liter, except as indicated; °C, degrees Celsius; µS/cm, micromsiemens per centimeter at 25 degrees Celsius. Water filtered through 0.45-micrometer filter before analysis except where indicated. <, less than.]

Glynn and others (1991) simulated chemical reactions and transport in the subsurface with the model PHREEQM and found a tendency for the predicted pH to be higher than pH measured in the field. Exsolution of CO₂ will be measured at the leading edge of the acidic part of the plume to determine the contribution of this process to the pH and redox balance and ultimately to contaminant transport. Preliminary gas measurements of CO₂ in the unsaturated-zone at site 500 indicate a diffusion-induced fractionation of 4.4 per mil for carbon ¹³C over a distance of 2.38 m between neutralized water at the water table and the land surface-the fractionation expected for a steady-state upward flux of CO₂ (Glynn and Busenberg, 1996). Pierre D. Glynn, (U.S. Geological Survey, National Research Program, Reston, Virginia), plans to measure exsolution and estimate the flux of CO₂ at or near the leading edge of the acidic part of the plume.

The testing, verification, and application of coupled geochemical-transport models also will help answer some of the remaining questions about subsurface-contaminant transport. General testing of reactive-transport models and the use of sensitivity analyses to define the most important reactions will be continued by Glynn, in cooperation with Peter Engesgaard (Technical University, Denmark) and Kenneth L. Kipp (U.S. Geological Survey, Denver, Colorado). James G. Brown, (U.S. Geological Survey, Tucson, Arizona), R.L. Bassett (University of Arizona, Tucson, Arizona), and Glynn plan to use a sensitivity- analysis type of approach to identify and determine the relative importance of chemical reactions, dilution, and transport that may control the remediation and chemical recovery of the regional aquifer. Brown, Bassett, and Glynn also plan to simulate transient, reactive transport in two dimensions, accounting for redox, adsorption, dissolution-precipitation, complexation, and hydrolysis.

Ground-Water Hydrology

Adequate simulation of reactive-solute transport through the subsurface depends on the accuracy of estimating recharge to the aquifer and subsurface-hydrological transport. D.R. Pool (U.S. Geological Survey, written commun., 1993) estimated changes in storage and specific yield in the alluvium on the basis of observed changes in ground-water level and gravity, and plans similar measurements between Miami Wash and Pinal Creek. Average values of specific yield of sediments that were saturated by water-level rises are from 0.16 to 0.21. Larger specific-yield values were associated with sediments containing large percentages of sand and gravel. Small specific yields were associated with sediments finer than sand.

Pierre Glynn and Eurybiades Busenberg (U.S. Geological Survey, National Research Program, Reston, Virginia) measured chlorofluorocarbon (CFC) concentrations in ground water to estimate ground-water ages in the basin. Preliminary results using CFC-11 for age-dating indicate that ground water from a depth of about 18 to 27 m in alluvium upgradient from contamination sources was recharged in 1979 (Glynn and Busenberg, 1996). Planned additional measurements of CFC's may indicate if an assumption of steady-state flow for reactive-transport modeling is valid. Travel-time information derived from these measurements, combined with the results from geochemical modeling, may provide estimates of chemical reaction-rates at the site.

Metal Transport in Shallow Ground Water and Streamflow

The acidic ground-water plume has moved downgradient at a rate of about 0.25 km/yr over the past several decades (Eychaner, 1991a). Higher-than-normal rainfall in 1991-93 increased ground-water levels by as much as 15 m and moved the headwaters of perennial flow upvalley. In early 1993, acidic-contaminated ground water in Miami Wash rose temporarily to the ground surface and discharged directly to streamflow. The headwater of perennial streamflow has since reestablished itself 1 km upstream from the location shown in figures 1 and 3. Increased hydraulic gradients and shortened ground-water flow paths may cause the acidic core of the plume to reach perennial surface water sooner than expected, depending on the transport of uncontaminated recharge water and the degree of mixing of uncontaminated recharge water with the contaminated water in the plume. Chemical mass-balance models and ground-water-flow modeling, where needed, will be used by Harvey and others to investigate the consequences of enhanced interaction between ground water and surface water on mass-flow rates of contaminants out of the basin.

Neutralized contaminated ground water with large Mn concentrations (table 1, well 503) is being discharged to the perennial reach of Pinal Creek. High pCO₂ and low pO₂ of contaminated ground water is modulated by gas exchange with the atmosphere after ground water discharges to the stream. Resulting increases in pH and pe favor Mn oxidation (Eychaner, 1991b). Concentrations of colloidal Fe and Mn presently in streamflow are low compared to total concentrations in solution (Harvey and Fuller, 1996). The absence of high Mn colloid concentrations suggests that Mn is partitioned to solid phases mostly on surfaces of streambed solids. Colloidal transport of Fe, Mn, and other trace metals appears in general to be only a minor mechanism of metal transport during baseflow in Pinal Creek. The role of Fe colloids in trace-metal transport will likely increase if the acidic core of the contaminated plume reaches Pinal Creek.

The character and chemical composition of manganese precipitates on the streambed at Pinal Creek vary considerably (Lind and Hem, 1993). J.D. Hem and C.J. Lind (U.S. Geological Survey, Menlo Park, California) are continuing their studies of chemical controls on manganese mineral formation at the site. Mn carbonate precipitation on streambed sediments appears to be favored initially in upstream reaches; Mn oxide precipitation becomes more important as pH and availability of dissolved oxygen both increase farther downstream (Hem and Lind, 1996). Fe and Mn oxides and carbonate minerals on the streambed scavenge other dissolved trace metals, including nickel (Ni), copper (Cu), and cobalt (Co), from surface flow. Trace metals have an decreasing affinity for Fe oxides, Mn oxides, and carbonate minerals, in decreasing order (Lind and Anderson, 1992). The greater abundance of Mn solid phases compared to Fe solid phases suggests that trace metal partitioning may be controlled by reactions with Mn phases. Chris Fuller (U.S. Geological Survey, Menlo Park, California) is investigating the relative roles of trace-metal coprecipitation and adsorption on Mn phases on the streambed as mechanisms for attenuation of dissolved Ni, Cu, and Co in streamflow (Fuller, 1996). The goals are to determine dominant chemical mechanisms of trace-metal partitioning to solid phases and to estimate partitioning coefficients for improved hydrochemical modeling.

Harvey and others (1996) showed that significant partitioning of Mn and Fe occurs between solid and aqueous phases in shallow ground water before ground water is discharged into Pinal Creek. A gain in Fe and loss of Mn from solution occurs in ground water that discharges laterally though the stream-channel banks at Pinal Creek; in contrast, ground water that discharges vertically through the creek bottom is unaltered by chemical reaction. A coupled mass balance for 500 m of stream and the surrounding aquifer in November 1992 demonstrated the importance of reactions in shallow ground water on downstream water chemistry in Pinal Creek.

Water exchange between streamflow and streamwater-filled flow paths in streambed sediments was investigated at Pinal Creek by a bromide-tracer experiment. Exchange between streamflow and ground water affects the contact area and contact times of streamflow with streambed solids. Rates of Mn oxidation and trace-metal sorption on streambed surfaces, therefore, may be sensitive to surface water-ground water hydrological exchange. Preliminary data are being used by Harvey, Fuller, and Conklin (University of Arizona, Tucson, Arizona) to design future experiments at Pinal Creek. Aspects of identifying the structure of improved hydrochemical models for Pinal Creek and of estimating model parameters and associated uncertainties are outlined in Wagner and Harvey (1996).

SUMMARY

Acidic drainage from copper mining has contaminated the regional alluvial aquifer in the Pinal Creek basin. The acidic contamination is neutralized along the flowpath by reactions with aquifer materials and native ground water. Neutralized water discharges in a perennial reach of Pinal Creek; as the water equilibrates with the atmosphere, pH increases and Mn minerals precipitate on the streambed.

Ongoing research at the site is focused on ground-water geochemistry, hydrology, and solute transport and interactions between ground water and perennial streamflow. Field and laboratory experiments will attempt to identify the reduction(s) that balance the oxidation of Fe in the aquifer. Two-dimensional reactive-transport simulations of plume evolution and remediation will identify the important processes and reactions that produced the observed conditions in the plume.

Measurements of changes in gravity and levels of ground water at project wells may provide refined estimates of aquifer specific yield for use in reactive-transport simulations. Estimates of ground-water ages obtained by CFC age-dating techniques may provide improved estimates of ground-water traveltimes and, when combined with results from geochemical modeling, provide estimates of geochemical-reaction rates at the site.

Field, laboratory, and modeling studies of contaminant transport are being conducted to improve understanding of the controls on metal transport in near-stream ground water and in streamflow. Previous work showed that Mn precipitates are variable and complex, and that precipitation is generally controlled by CO₂ outgassing and reareation, which affects pH and the solubility of Mn solids. Mechanisms of trace-metal scavenging by Mn carbonates, Mn oxides, and Fe oxides are being investigated, and partitioning coefficients are being estimated to improve hydrogeochemical modeling. Tracer experimentation and a coupled aquifer-stream mass-balance approach are being used to determine transport components and to constrain net effective rates of solid-aqueous partitioning of Mn, Fe, Ni, Cu and Co in the system.

REFERENCES CITED

Eychaner, J.H., 1989,

Movement of inorganic contaminants in acidic water near Globe, Arizona, in Mallard, G.E., and Ragone, S.E., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Phoenix, Arizona, September 26-30, 1988: U.S. Geological Survey Water-Resources Investigations Report 88-4220, p. 567-575.

Eychaner, J.H., 1991a,

The Globe, Arizona, research site-contaminants related to copper mining in a hydrologically integrated environment, in Mallard, G.E., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Monterey, California, March 11-15, 1991: U.S. Geological Survey Water-Resources Investigations Report 91-4034, p. 439-447.

Eychaner, J.H., 1991b,

Solute transport in perennial streamflow at Pinal Creek, Arizona, in Mallard, G.E., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Monterey, California, March 11-15, 1991: U.S. Geological Survey Water-Resources Investigations Report 91-4034, p. 481-485.

Fuller, C.C., 1996,

Expermental studies of trace-metal partitioning in perennial reaches of Pinal Creek, Arizona, in Morganwalp, D. W., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Colorado Springs, Colorado, September 20-24, 1993: U.S. Geological Survey Water-Resources Investigations Report 94-4015.

Glynn, P.D., 1991,

Effect of impurities in gypsum on contaminant transport at Pinal Creek, Arizona, in Mallard, G.E., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Monterey, California, March 11-15, 1991: U.S. Geological Survey Water-Resources Investigations Report 91-4034, p. 466-474.

Glynn, P.D., and Busenberg, E., 1996,

Unsaturated zone investigations and chlorofluorocarbon dating of ground waters in the Pinal Creek basin, Arizona, in Morganwalp, D. W., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Colorado Springs, Colorado, September 20-24, 1993: U.S. Geological Survey Water-Resources Investigations Report 94-4015.

Glynn, P.D., Engesgaard, Peter, and Kipp, K.L., 1991,

Use and limitations of two computer codes for simulating geochemical mass transport at the Pinal Creek toxic-waste site, in Mallard, G.E., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Monterey, California, March 11-15, 1991: U.S. Geological Survey Water-Resources Investigations Report 91-4034, p. 454-460.

Harvey, J.W., and Fuller, C.C., 1996,

Association of metal contaminants with colloidal and suspended particulate material in shallow ground water and surface water at Pinal Creek, Arizona, in Morganwalp, D. W., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Colorado Springs, Colorado, September 20-24, 1993: U.S. Geological Survey Water-Resources Investigations Report 94-4015.

Harvey, J.W., and Fuller, C.C., and Wagner, B.J., 1996,

Interactions between shallow ground water and surface water that affect metal contaminant transport in Pinal Creek basin, Arizona, in Morganwalp, D. W., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Colorado Springs, Colorado, September 20-24, 1993: U.S. Geological Survey Water-Resources Investigations Report 94-4015.

Hem, J.D., and Lind, C.J., 1996,

Chemical processes in manganese oxide and carbonate precipitation in Pinal Creek, Arizona, in Morganwalp, D. W., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Colorado Springs, Colorado, September 20-24, 1993: U.S. Geological Survey Water-Resources Investigations Report 94-4015.

Lind, C.J., and Anderson, L.D., 1992,

Trace metal scavenging by precipitating Mn and Fe oxides, in Kharaka, Y.F., and Maest, A.S., eds., Proceedings of the 7th International Symposium on Water-Rock Interaction, July 13-18, 1992: WRI-7, p. 397-402.

Lind, C.J., and Hem, J.D., 1993,

Manganese minerals and associated fine particulates in the streambed of Pinal Creek, Arizona-A mining-related acid drainage problem: Applied Geochemistry, v. 8, no. 1, p. 67-80.

Longsworth, S.A., and Taylor, A.M., 1992,

Chemical, geologic, and hydrologic data from the study of acidic contamination in the Miami Wash-Pinal Creek area, Arizona, water years 1990-91: U.S. Geological Survey Open-File Report 92-468, 59 p.

Peterson, N.P., 1962,

Geology and ore deposits of the Globe-Miami district, Arizona: U.S. Geological Survey Professional Paper 342, 151 p.

Wagner, B.J., Harvey, J.W., 1996,

Solute transport parameter estimation for an injection experiment at Pinal creek, Arizona in Morganwalp, D. W., and Aronson, D.A., eds., U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Colorado Springs, Colorado, September 20-24, 1993: U.S. Geological Survey Water-Resources Investigations Report 94-4015.