USGS Mine Drainage Newsletter

Application of the Microbial and Spectral Reflectance Technique
(MAST) to the Identification of Acid Mine Drainage at Contrary
Creek, Louisa County, Virginia
by Eleanora Robbins (1), John Anderson (2), Marta Flohr (1), Gordon Nord, Jr. (1), Melvin Podwysocki (1), Byron Prugh, Jr. (3), Mark Stanton (4), and Palmer Sweet (5).

(1) U.S. Geological Survey, Reston, VA; (2) U.S. Army Topographic Engineering Center, Alexandria, VA; (3) U.S. Geological Survey, Richmond, VA; (4) U.S. Geological Survey, Denver, CO; (5) Virginia Division of Mineral Resources, Charlottesville, VA

Introduction

Figure 1. Locality map showing location of Contrary Creek and neutral tributary (drain). Location of the Sulphur Mine is noted. (From Anderson and others, 1995).

Contrary Creek in Louisa County (Fig. 1), arguably the worst acid mine drainage (AMD) site in Virginia, receives drainage from five abandoned pyrite mines including the Sulphur, Boyd Smith, and Arminius Mines. The pyrite, which was mined for production of sulfuric acid, occurs in lenses up to 50 feet thick and several hundred feet long in the Cambrian Chopawamsic Formation, a pile of metamorphosed mafic to felsic volcanic rocks (Pavlides, 1989; Rankin, 1994). Iron gossan, copper, and pyrite were mined at intervals between 1834 and 1922 (Poole, 1973; Sweet and others, 1989). At the Sulphur Mine, underground mine workings include several drifts that undercut Contrary Creek. Although Contrary Creek is noted for gold panning, only data on silver production have been published (Sweet, 1976).

                             Table 1
           Dissolved Metals in Contrary Creek, Virginia

               Metal               Concentration
                                   (micrograms per liter)
                Cd                           5
                Cu                         320
                Fe (total)              11,000
                Pb                           8
                Mn (total)                 870
                Ni                           6
                Zn                       1,900


Sample collected November 6, 1989 (Prugh and others, 1990).

Water flowing through fractures in sulfide-rich country rocks and mine tailings has led to a continuous discharge of acidic drainage into Contrary Creek since mining began. Attempts to remediate the AMD have not been successful. Typical pH values in the perennial creek are about 3.5 (Bell and others, 1990), with a high of 4.8 in November 1989 (early winter) and a low of 2.9 in September 1991 (late summer). Dissolved metals in the creek were measured in November 1989 (Table 1). This same creek is fed by near-neutral ground-water seeps and tributaries. The pH of one such tributary ranges from 6.0 to 6.5 (Anderson and others, 1995).

The continuing generation of AMD at Contrary Creek presents an opportunity for a diverse and ever-expanding team of scientists, the "Red Slime Team", to bring the knowledge, tools, and techniques of their respective disciplines to bear on the problem. The team consists of scientists and engineers from the Geologic and Water Resources Divisions of the U.S. Geological Survey, and also from the U.S. Army Corps of Engineers-Topographic Engineering Center and the Virginia Division of Mineral Resources of the Virginia Department of Mines, Minerals and Energy, who will jointly design the study and collect and analyze the data. The results of the study will be disseminated to team members from the U.S. Bureau of Mines (and successor agencies), Office of Surface Mining, U.S. Environmental Protection Agency, National Park Service, and Pennsylvania Division of Environmental Resources, who will apply the findings and provide independent tests of hypotheses.

Hypotheses presently being tested by the team are: (1) that the pH of iron-rich streams can be approximated using remote sensing methods to distinguish acidic from near-neutral chemistries (Robbins and others, 1995) and (2) that the specific conductance of acidic waters can be estimated by spectral measurements. We have applied the acronym "MAST" (Microbial And Spectral reflectance Technique) to the remote sensing methodologies being tested. The hypotheses are based on observations in the visible and reflective-infrared parts of the spectrum that suggest that the iron precipitates of neutral iron bacteria have different spectral responses compared to those of acid iron-oxidizing bacteria. In essence, the iron bacteria that thrive in near-neutral-pH waters actively precipitate iron oxide minerals that appear red or red-orange in the visible spectrum, whereas the iron-oxidizing bacteria that thrive in acid pH water actively precipitate iron oxide minerals that appear yellow or yellow-orange ("yellow boy").

Figure 2. In situ averaged field spectra for precipitates in acid (solid line) and near-neutral (dashed line) streams. Both acid and near-neutral bacterial precipitates have distinct spectral properties in the visible and near reflected infrared (NIR) parts of the spectrum which are temporally and spatially consistent. Both the acid and near-neutral bacterial precipitates have intense ferric-iron absorption features in the ultraviolet (UV) part of the spectrum (wavelengths < 400 nm), causing a diminished reflectance in the blue and green parts of the spectrum. Both groups also have additional weaker absorption features in the 800-900 nm (NIR) region. Both the UV and the NIR features are due to electronic absorption bands related to the valence shell states in the ferric iron atom. However, both the amplitude of the reflectances in the two groups as well as the exact positions of their absorption features are sufficient to distinguish and to identify the two different groups of iron precipitates. (After Anderson and others, 1995).

Microbial Sampling - Methods

Microbial Sampling - Results

Figure 3. A plot of spectral reflectance versus wavelength for wet and dry samples of ferric-iron bacterial flocs. Spectral curves A and B show the acid and neutral bacterial flocs, in their native water media. Measurements were made on water-suspended flocs within a 4-mm-deep i sample holder. The flatness of these spectral curves at wavelengths greater than 1400 nm is due to intense absorption by liquid water in this spectral region. The absorption features at approximately 950 and 1200 nm are due to less intense water absorption. Note that the acid floc is brighter than the neutral one, a characteristic verified by spectra taken above the floc-laden i water bodies in the field. The falloffs in reflectance from the 750 nm peak towards 500 micrometer and 900 nm are characteristic of ferric iron absorption bands centered in the 800-900 nm (NIR) and ultraviolet regions (< 400 nm). The falloff in reflectance from 750 nm towards 900 nm is more intense for the acid flocs. Spectral curves C and D represent the same flocs, but air dried in their sample holders at room temperature. The absorption features at 1400 and 1900 nm are due to hydroxyl anion and water, respectively. Note that the acid floc has an overall greater reflectance, especially in the NIR, as well as a more intense absorption band centered in the NIR region compared to the neutral floc. Spectral curves for dried flocs normally are brighter than their wet equivalents. However, because the flocs have a tendency to fragment into irregular polygons with curled up edges (see also Plate 1, fig. 7) while drying, the spectrally neutral flat black paint of the sample holder is exposed, thus diminishing the overall brightness of the air dried samples.

Spectral Measurements - Methods

Measurements of the distinct spectra also have been successfully repeated with a Digital Multispectral Video system using a light fixed-wing aircraft flying at 5500 ft above ground surface. The aerial experiments indicate that the technique probably will be useful for identifying point-source AMD incursions into streams using more sophisticated imaging spectrometer systems.

Laboratory spectral reflectance measurements were taken for wavelengths from 400 to 2500 nm using a Beckman 5240 spectrometer. The most useful data reside in the 650 to 1300 nm near-infrared (NIR) window.

Figure 4. Atomic models of two-line ferrihydrite (a and b) and goethite (c). The 8-sided polygons are FeO6- octahedra representing 6 oxygens at the apices and one Fe+3 atom in the center. The small atoms bonded to the corners of the octahedra are hydrogen. Two different ways of combining FeO6- octahedra at the edges and corners are shown in a and b. Either of these combinations have been suggested to represent the structure of two-line ferrihydrite (Waychunas and others, 1993). The small size of the crystals gives rise to very broad X-ray diffraction maxima. After heating the ferrihydrite for 10 hours at 90 degrees C, one broad maximum characteristic of the mineral goethite appears in the X-ray diffraction pattern. The atomic model in 4c is goethite, FeOOH, and the size of the unit cell is shown in the center of the model. One can see that the combinations in 4a and 4b are parts of the goethite structure. The two lattice planes, {110}, indicated in the goethite model contain the greatest density of Fe atoms. X-ray diffraction from these planes gives rise to the characteristic broad maximum at 4.18 Angstroms.

Spectral Measurements - Results

The field spectral measurements clearly differentiate between the higher reflectance of the more yellow colors in acid waters and the lower reflectance of more red colors in neutral waters (Fig. 2). Laboratory results corroborate field measurements that show wet acid flocs brighter than neutral flocs in the 500 to 950 nm range (cf. figs. 2 and 3). The ferric iron absorption band centered in the 800 to 900 nm region is more intense in the acid than in the neutral floc. Given the similar particulate sizes of the acid and neutral flocs, the brightness difference may relate to greater long-range order in the ferrihydrite of the acid flocs.

Mineralogical Analysis

Plate 1.

Acid loose yellow floc

1. short rods (sr)


Neutral loose red floc


2. L. ochracea (Lo) (straws) colonized by Siderocapsa sp.
(Si) (globular)
3. Fungal hypha (septate)
4. diatom
5. fecal pellet of zooplankton
6. Gallionella ferruginea (G) (braid)


Acid oil-like film

7. short rods (sr) (note cracks in dried film)


Neutral oil-like film

8. short rods (sr), long rods (lr), rod chain (rc), and
holdfasts of L. discophora (Ld)


Acid settle

9. short rods (sr)


Neutral settle

10. diatom
11. short rods (sr), long rods (lr), rod chains (rc), and
holdfasts of L. discophora (Ld)

Discussion and Conclusions

The techniques described here are intended for application to perennial streams, but application of a similar remote-sensing approach to ephemeral streams in arid environments may also prove to be productive.

REFERENCES

Anderson, J.E., Robbins, E.I., and Podwysocki, M.H., 1995 (submitted), Evaluating spectral reflectance of bacterial precipitates in acid-mine and neutral drainage to monitor water quality at Contrary Creek, Mineral, Virginia: A possible technique for remotely sensing acidified watersheds: Remote Sensing of Environment, 20 p.

Bell, P.E., Herlihy, A.T., and Mills, A.L., 1990, Establishment of anaerobic, reducing conditions in lake sediment after deposition of acid, aerobic sediment by a major storm: Biogeochemistry, v. 9, p. 99-116.

Chukhrov, F.V., Zvyagin, B.B., Gorshkov, A.I., Yermilova, L.P., and Balashova, V.V., 1974, Ferrihydrite: International Geology Reviews, v. 16, p. 1131-1143.

Ferris, F.G., Tazaki, K., and Fyfe, W.S., 1989, Iron oxides in acid mine drainage environments and their association with bacteria: Chemical Geology, v. 74, p. 321-330.

Pavlides, Louis, 1989, Early Paleozoic composite melange terrane, central Appalachian Piedmont, Virginia and Maryland, its origin and tectonic history, in Melanges and Olistromes of the U.S. Appalachians: Geological Society of America, Special Paper 228, p. 135-193.

Poole, J.L., 1973, Iron sulfide mines in Virginia: Virginia Minerals, v. 19, p. 29-33.

Prugh, B.J., Jr., Easton, F.J., and Belval, D.L., 1990, Water Resources Data, Virginia, Water Year 1990. U.S. Geological Survey Water-Data Report VA-90-1.

Rankin, D.W., 1994, Continental margin of the eastern United States: Past and present, in Speed, R.C. (ed.), Phanerozoic evolution of North American continent-ocean transitions: Geological Society of America, Decade of North American Geology, Continent-Ocean Transect Volume, p. 129-218.

Robbins, E.I., and Norden, A.W., 1995, Microbial oxidation of iron and manganese in wetlands and creeks of Maryland, Virginia, Delaware, and Washington. D.C., in Chiang, S.H., (ed.), Pittsburgh Coal Conference Proceedings: Coal -- Energy and the Environment, v. 2, p. 1154-1159

Robbins, E.I., Anderson, J.E., Podwysocki, M.H., Edenborn, H.M., Eggleston, J., Growitz, D., Kleinmann, R.L., Norden, A.W., Passmore, M., Prugh, B.J., Jr., Stanton, M.R., Sweet, P.C., and Vandyke, T.M., 1995, Red slime, oily films, and black rocks--Sensing the iron bacteria that make neutral streams look acid, in Carter, L. (ed.), USGS McKelvey Forum on Energy and the Environment: U.S. Geological Survey Circular 1108, p. 109-111.

Satterwhite, M.B., and Henley, J.P., 1990, Hyperspectral signatures (400 to 2500 nm) of vegetation, minerals, soils, rocks, and cultural features: Laboratory and field Measurements: U.S. Army Corps of Engineers Engineer Topographic Laboratories Report ETL-0573, p. 21-24.

Sweet, P.C., 1976, Silver in Virginia: Virginia Division of Mineral Resources Virginia Minerals, v. 22, p. 9-11.

Sweet, P.C., Good, R.S., Lovett, J.D., Campbell, E.V.M., Wilkes, G.P., and Meyers, L.L., 1989, Copper, lead, and zinc resources in Virginia: Virginia Division of Mineral Resources, Publication 93, 185 p.

Waychunas, G.A., Rea, B.A., Fuller, C.C., and Davis, J.A., 1993, Surface chemistry of ferrihydrite: Part 1. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate: Geochimica and Cosmochimica Acta, v. 57, p. 2251-2269.