Water Quality Changes in an Acid Mine Drainage
Stream Over a 25-Year Period

¹Jason Stewart was a graduate research assistant at the time of the research, and Jeff Skousen is Professor of Soil Science, Division of Plant and Soil Sciences, West Virginia University, Morgantown, WV26506-6108, USA.

Keywords: Clean Water Act (CWA), coal mining, Deckers Creek, fecal coliforms, land reclamation, Surface Mining Control and Reclamation Act (SMCRA), water pollution, watershed

Abstract

    The Deckers Creek watershed in northern West Virginia, containing a land area of 166 km² (63 mi²) has a long history of industrial development and its attendant environmental abuses from both land and water pollution practices. The passage of laws regulating discharges into the waters of the United States, enacted during the 1970s, has had a measureable impact on streams and rivers water quality, and is demonstrated in Deckers Creek. The water in Deckers Creek was sampled in 1974 and re-sampled in 1999 to determine water quality changes over this 25-year period. Twenty-nine sampling points were located along the main stem of the creek and water samples were analyzed for pH, acidity, alkalinity, iron, and calcium at both times. Aluminum, manganese and fecal coliform (FC) densities were measured only in 1999. Water at all sampling points showed lower acidity and metal contents in 1999 compared to 1974. Water pH increased at the mouth from 4.8 in 1974 to 6.3 in 1999. Acidity was reduced an average of 62% across all sites and iron concentrations were decreased 53%. However, one major untreated point source of water from an abandoned underground mining complex continues to degrade the quality of the creek in its lower stretches. In the upper stretches, the water quality in Deckers Creek has improved due to decreased coal mining activities, reclamation of abandoned and recently-permitted mined sites, and healing over time of past land use scars (timbering and mining) due to natural processes. More time and additional reclamation projects will continue to enhance the water quality in the creek. Improved water chemistry in the majority of the creek, however, allows detection of biological contamination from sewage inputs and presents a new water quality challenge in the watershed.

    The Clean Water Act (CWA) was passed in 1977, which included amendments of the previously passed Federal Water Pollution Control Act of 1972 (Arbuckle et al. 1993). The purpose of the CWA was "to restore and maintain the chemical, physical, and biological integrity of the nation’s waters." To achieve this purpose, the CWA established a permitting system (National Pollutant Discharge Elimination System, NPDES) which regulates companies or individuals discharging water from point sources into rivers or lakes of the United States. The NPDES program established effluent standards for these discharges and required compliance from those issued these permits. The NPDES program was intended to maintain the quality of our nation’s waterways and to restore water bodies that had been previously degraded. As a result, all chemical, manufacturing, energy, and mineral extraction companies have been required to obtain NPDES permits with effluent standards for discharged water since 1977.

    The Surface Mining Control and Reclamation Act (SMCRA) became law in 1977. SMCRA established mining performance and land reclamation standards, including the assurance that "surface coal mining operations are so conducted as to protect the environment" and to promote the "reclamation of mined areas left without reclamation prior to" the establishment of SMCRA. It also established a permitting system where companies must comply with those mining and reclamation standards listed in the permit. However, SMCRA recognized the CWA, and water discharge points from coal mines were regulated under the CWA. Therefore, coal mining companies must get surface mining permits under SMCRA and also NPDES water discharge permits under CWA.

    The effects of these two laws on water quality improvements often have not been recognized because knowledge and memory of changes in land and water resources over decades of time are forgotten or not passed on to succeeding generations. The Deckers Creek watershed has a long history of industrial development and its attendant environmental abuses from both land and water pollution practices, and is used as a case study to evaluate the effects of these two federal laws on improvements in water quality over time.

    The Deckers Creek watershed lies within the drainage basin of the Monongahela River in western Preston and eastern Monongalia Counties in northern West Virginia (Figure 1). The stream is approximately 38.1 km (23.6 miles) in length, originating in Preston County at an elevation of 737 m (2,427 ft). The creek flows southeast for a short distance, then it loops to the north receiving water from Kanes and Dillan Creek tributaries. Relief is very gradual from the headwaters to Masontown before becoming steep for a long stretch flowing northwest from Masontown to Dellslow. After Dellslow, the creek again flows over gently sloping terrain until reaching the Monongahela River at an elevation of 237 m (793 ft) (Teti 1974).

            Figure 1

    Most of the exposed geology is part of the Pennsylvanian Period, namely the Monongahela, Conemaugh, and Pottsville Groups and the Allegheny Formation (Natural Resources Conservation Service (NRCS) 2000). The dominant rock types include sandstone, siltstone, shale, limestone, and coal. Major soils in the area include the Gilpin (fine-loamy, mixed, semiactive, mesic, Typic Hapludults) and Dekalb (loamy-skeletal, silicious, active, mesic Typic Dystrochrepts) series on the uplands and the Atkins (fine-loamy, mixed, active, acid, mesic Fluvaquentic Endoaquepts) and Pope (coarse-loamy, mixed, active, mesic Fluventic Dystrochrepts) series in the bottomlands (NRCS 2000).

    The land surface is part of the Allegheny Mountain section, which includes the highest elevations of West Virginia, and is generally composed of mountain ranges oriented in a northeast-southwest direction. The vegetation is classified as the Northern Hardwood cover type, with 70% of the land surface in the watershed covered by the oak-hickory (Quercus - Carya) forest type (Strausbaugh and Core 1977). Farmland comprises 15% of the watershed, with nearly all of the active farming occurring in the upper portions of the watershed in Preston County. Urban land, found in and around the cities of Masontown and Morgantown, makes up roughly 10% of the area. Mined land accounts for the remaining 5% (NRCS 2000).

    The area was first settled in 1772 along Deckers Creek near Morgantown. By 1798 one of the first iron works west of the Blue Ridge Mountains was located at Rock Forge beside a steep portion of the creek, just 4 km east of Morgantown. The population of Morgantown and surrounding areas grew slowly through the 1800s as these small industries gradually developed. Timbering operations clear-cut the northern West Virginia forest in 1870-1880. The second growth forest was again cut in the 1900-1920 period, and again in the 1930s to 1940s (Clarkson 1964; Hicks 1998). During the timber boom, observers wrote "the great mountain slopes and forest had been ruinously detimbered, the soil on the hillsides had eroded and washed down, and the surface mines had been deserted leaving raw scars…" (Eller 1985).

    Numerous coal mines opened throughout both Monongalia and Preston Counties within the watershed in the early 1900s (Moreland 1985). The first commercial coal mine in the Deckers Creek watershed was located at Richard (5 km east of Morgantown) in 1903, when the West Virginia Coal Company opened an underground drift mine into the Upper Freeport coal along the north bank of Deckers Creek. By the end of that first year, 52 miners and laborers had produced enough coal to yield over 6,350 metric tons (7,000 tons) of coke by burning the coal in the 75 original beehive coke ovens (Paul 1904). The operation grew so that by 1919, the mine workers were delivering coal to the surface at a rate of 60,820 metric tons (67,000 tons) per year, yielding 41,200 metric tons (45,377 tons) of coke per year. Mining continued at the Richard Mine until the 1950’s, when production ceased.

    When the mine was abandoned, water accumulated in the underground mine voids and seeps eventually appeared at the land surface at many points, with the major water outflow from the mine coming at the portal and entering Deckers Creek. The reason for poor quality water to drain from the Richard Mine is due to the coal itself. Deckers Creek flows through a high-sulfur coal region in West Virginia, where the coal and associated strata contain high levels of pyrite. When exposed to water and oxygen during mining, pyrite undergoes a series of chemical reactions, ultimately resulting in the release of sulfate, proton acidity, and iron (Geidel and Caruccio 2000). As elevated levels of iron are introduced into natural waters, the iron is oxidized and hydrolyzed, thereby forming iron hydroxides and acidity (Rosseland et al. 1992). The iron hydroxides precipitate out of solution and coat stream rocks and sediments, causing a distinct orange color staining (Skousen et al. 2000a; Younger 1998). Low pH conditions in the water accelerate weathering and the dissolution of silicate and other rock minerals, thereby causing the release of other elements such as aluminum and manganese into the water (Kittrick et al. 1982).

    Acidity and metals, especially aluminum, have detrimental impacts on aquatic organisms (Gray 1995; Neville and Campbell 1988; Sparling and Rowe 1996). In a study of benthic macroinvertebrates in Deckers Creek, Mains et al. (1999) found that most of the stream contains a very low diversity of organisms, and that the total number of organisms is also very low. The most heavily damaged section of the stream was below the Richard Mine discharge, where the creek rated "very poor," and no fish were reported.

    The addition of limestone to acid mine drainage streams neutralizes the acidity and causes metals to precipitate from solution. A large limestone input to Deckers Creek comes from an active limestone quarry, Greer Limestone, which stores alkaline tailings along a 500-m (1640-ft) section of floodplain about halfway between the headwaters and the mouth at Morgantown (Figure 1). The alkaline input at Greer is thought to be responsible for the improved quality of Deckers Creek from Greer down to the inflow of untreated acid mine drainage at the Richard Mine discharge.

    Through the early to mid 1900s, mining became important to the economy in northern West Virginia, bringing more people and other industries to the Deckers Creek watershed. The increased industrial activity led to the addition of higher amounts of acid mine drainage, chemicals, and sediment to the creek. Many residents along the creek used Deckers Creek for trash and sewage disposal. High levels of pollutants from soap, oil, sediment, as well as nitrogen, phosphorus, and pathogens in sewage were introduced to the stream. Eventually the pollution levels exceeded the attenuation capacity of the creek, resulting in widespread deterioration of water quality. By 1935, recreational use of Deckers Creek had stopped and the State Health Department at the time identified the creek as a possible health hazard due to its high levels of sewage contamination. A sewage collection and treatment system was completed in the Morgantown area by 1962, but acid mine drainage and sewage continued to enter the creek at various uncontrolled points along its stretch (NRCS 2000).

    Fecal coliforms (FC) and fecal streptococci (FS) are two of the more commonly isolated organisms for identifying sewage inputs into streams. These organisms are found in the gastrointestinal tract of mammals. So the extent to which natural waters have been impacted by fecal matter in sewage from mammals may be determined through the isolation and enumeration of FS and FC (Howell et al. 1995; Thelin and Gifford 1983; Young and Thackston 1999).

    In 1974, James Teti, a graduate student at West Virginia University, conducted a water quality study of Deckers Creek (Teti 1974) by sampling and analyzing the water at 29 points along the creek during a six-month period. The objectives of this study were to revisit the original sites sampled by Teti, collect water samples over an 18-month period and, after water analyses, evaluate the changes in water quality between 1974 and 1999.

1974 Sampling Procedures

    Water samples were collected in plastic, one-liter bags at 29 sample sites along Deckers Creek (Figure 1) on eight occasions between January and June of 1974. Teti (1974) reports withdrawing each sample from a flowing part of the stream, filling the sample containers completely, placing the water samples on ice, and analyzing the water in the laboratory for pH, acidity, alkalinity, iron, calcium, and magnesium. Teti did not measure flow when he collected water samples during the 1974 sampling period.

1999-2000 Sampling Procedures

    Beginning in March 1999, water samples were collected at 29 sites along Deckers Creek, which corresponded to sites originally sampled by Teti. A YSI 3500 Water Quality Meter (Yellow Springs Instrument Co., Yellow Springs, OH) was used to measure temperature, electrical conductivity, and pH in the field. Two water samples were collected at each site. The first sample was collected in a 250-ml plastic bottle and was neither filtered, nor acidified. A second sample of 20 ml was filtered with a 0.45 um filter and acidified to pH 2.0 with 0.5 ml of concentrated hydrochloric acid. Both samples were placed on ice and transported to the laboratory. The first sample was analyzed for pH, acidity, and alkalinity using a TitraLab autotitrator (Radiometer Co., Copenhagen, Denmark), while the second sample was analyzed for total iron, aluminum, manganese, calcium, and several other elements using an Inductively Coupled Spectrophotometer, Plasma 400 (Perkin Elmer, Norwalk, CT). All titrations were performed within 6 hr of sample collection.

    Flow determinations were made at the same time water samples were taken. At the beginning of the study at each sample point, the cross-section of the stream channel was mapped and a reference point was marked for subsequent depth measurements. At later sampling times, the depth of the water was measured at this reference point and used to calculate the vertical area of water flow at the stream cross-section. Water velocity was measured using a Global Water Flow Probe FP101 (Global Water, Gold River, CA), which computed an average velocity over a ten-second period. Multiplying the cross-sectional area of the water in the stream by velocity provided the estimate of flow at that sampling point.

    Another water sample was collected two or three days after chemistry samples were obtained each month and tested for the presence of FC and, for a few months, FS. A membrane filtration technique was used (Greenberg et al. 1985). Most of the 29 sites were sampled monthly from March to June 1999, until it was determined which sites warranted further monitoring. Eight representative sites were then sampled monthly beginning in June 1999. Samples were collected in one-liter plastic sterilized bottles that had been treated with 3 ml of sodium thiosulfate to bind any residual chlorine present in the creek water. Six volumes were analyzed (0.1, 1.0, 5.0, 10.0, 50.0, and 100.0 ml) to produce plates with between 20 and 60 countable colonies. Both the 0.1-ml and the 1.0-ml volumes of sample were combined with 12 ml of peptone buffer in order to pass a sufficient volume of fluid through the filters. The 5.0-ml sample had 8.0 ml of peptone added, the 10-ml sample had 4.0 ml of peptone added, while the 50- and 100-ml volumes had no peptone buffer added. The tests were done with 0.7-um filters instead of 0.45-um filters, as this larger size filter allows for improved recovery of organisms possibly damaged by acidic conditions in the stream (Bissonnette, 2000).

    Samples analyzed for FC were plated onto m-FC agar, and incubated at 44.5ºC ± 0.2°C for 24 hrs. The high incubation temperature and the selective properties of the media helped in restricting growth to FC bacteria. Individual colonies with a bright blue color were counted. Samples analyzed for FS were plated onto m-Enterococcus agar, and incubated at 35°C ± 0.5°C for 48 hours. The selective properties of the media were responsible for limiting plate growth to FS only. Individual colonies with a bright red color were counted. The plates were then examined to determine the number of colony forming units (CFU’s) per 100 ml.

    Statistical software (SAS) was used to perform an analysis of variance (ANOVA) on the 1974 and 1999-2000 flow and chemistry data sets. Since significant differences were found among flow categories, means were calculated for the various water quality parameters for high, medium, and low flow periods. Analysis of variance was also performed on the FC data using SAS statistical software. In order to compensate for the large variations in FC counts normally seen in microbiological studies, geometric means were analyzed. When working with a data set with large variations, geometric means give a better measure of centrality than do arithmetic means (Hunter et al. 1999).

Flow Data

    Since no flow data were collected during the 1974 study, we assessed the climatic conditions during 1974 and estimated stream flow by the use of rainfall data. If flows varied greatly between these two sampling times, any differences in water chemistry may be due to dilution or concentration effects rather than changes in baseline conditions. Table 1 lists the monthly rainfall data and the mean measured monthly flow at the mouth of Deckers Creek for several years. Using the monthly rainfall data and measured flows in 1947-1948 and 1999-2000, we estimated the mean monthly stream flow in 1974. The calculated mean annual value was 2.5 m³/sec, which corresponded well with the measured flows of other years and especially correlated well with the year 2000 flow. These estimated flows in 1974 were used to separate flow data into high, medium, and low flow categories. This estimate also indicated that the water quality data gathered by Teti in 1974 was not from an excessively wet or dry year, and that the 1974 data can be compared to our 1999 and 2000 data.

Chemistry

    Based on estimated and measured flows, the 1974 and the 1999-2000 monthly data were divided into low flow (June-October), medium flow (November-January), and high flow (February-May) categories. We found significant differences in all values among flow conditions, with the lowest pH and alkalinity, and highest acidity and metal concentrations occurring during low flow periods (Table 2). We chose to highlight the high flow values (n=7 in 1999-2000, Table 2) in this paper largely because the sample size for the 1974 high flow conditions was greater than the other flow categories (high flow: n=5; low flow: n=1; and medium flow: n=2).

1Values within columns with the same letter were not significantly different at p<0.05.

                                            NA=Not available.

Arbuckle, J., F. Brownell, D. Case, W. Halbieb, J. Jenson, S. Landfair, M. Miller, K. Nardi, A. Olney, D. Sarvedi, J. Spensley, D. Steinway, and T. Sullivan. 1993. Environmental law handbook. 12^th Ed. Gov. Insti., Inc. Rockville, MD.

Baudart, J., J. Grabulos, J.P. Barusseau, and P. Lebaron. 2000. Salomnella spp. and fecal coliform loads in coastal waters from a point vs. nonpoint source of pollution. J. Environ. Qual. 29:241-250.

Bissonnette, G. 2000. Personal communication. West Virginia University. Morgantown, WV.

Clarkston, R. 1964. Tumult on the mountains: lumbering in West Virginia, 1770-1920. McClain Printing Co., Parsons, WV.

Coyne, M.S., and J.M. Howell. 1994. The fecal coliform/fecal streptococci ratio (FC/FS) and water quality in the Bluegrass region of Kentucky. Soil Science News and Views. Vol. 15, No. 9. University of Kentucky, Lexington. KY.

Davenport, C.V., E.B. Sparrow, and R.C. Gorden. 1976. Fecal indicator bacteria persistence under natural conditions in an ice covered river. Appl. Environ. Microbiol. 32:527-536.

Edwards, D.R., M. S. Coyne, T. C. Daniel, P. F. Vendrell, J. F. Murdoch, and P.A. Moore, Jr. 1997. Indicator bacteria concentrations of two northwest Arkansas streams in relation to flow and season. American Society of Agricultural Engineers 40(1):103-109.

Eller, R. 1985. Land as commodity: industrialization of the Appalachian forests, 1880-1940. In: B. Buxton and M. Crutchfield (eds.), The Great Forest: An Appalachian Story. Boone, NC.

Farrell-Poe, K.L., A.Y. Ranijha, and S. Ramalingam. 1997. Bacterial contributions by rural municipalities in agricultural watersheds. American Society of Agricultural Engineers 40(1):97-101.

Geidel, G., and F. Caruccio. 2000. Geochemical factors affecting coal mine drainage quality. Chapter 5. In: R. Barnhisel, R. Darmody, and L. Daniels (eds.), Reclamation of Drastically Disturbed Lands. 2^nd Ed. Agronomy Monograph 41, American Society for Agronomy, Madison, WI.

Gray, N.F. 1995. Field assessment of acid mine drainage contamination in surface and ground water. Environ. Geology 27:358-361.

Greenberg, A.E., R.R. Trussell, and C.S. Clesceri. 1985. Standard methods for the examination of water and wastewater. American Public Health Association, American Water Works Association, American Public Health Association, Washington D.C.

Hicks, R. 1998. Ecology and management of central hardwood forests. John Wiley & Sons, Inc., New York.

Howell, J.M., M.S. Coyne, and P.L. Cornelius. 1996. Effect of sediment particle size and temperature on fecal bacteria mortality rates and the fecal coliform/fecal streptococci ratio. J. Environ. Qual. 25:1216-1220.

Howell, J.M., M.S. Coyne, and P.L. Cornelius. 1995. Fecal bacteria in agricultural waters of the Bluegrass region of Kentucky. J. Environ. Qual. 24:411-419.

Hunter, C., J. Perkins, J. Tranter, and J. Gunn. 1999. Agricultural land-use effects on the indicator bacterial quality of an upland stream in the Derbyshire Peak district in the U.K. Wat. Res. 33(17):3577-3586.

Jawson, M.D., L.F. Elliott, K.E. Saxton, and D.H. Fortier. 1982. The effect of cattle grazing on indicator bacteria in runoff from a Pacific northwest watershed. J. Environ. Qual. 11:621-627.

Kittrick, J., D. Fanning, and L. Hossner. 1982. Acid sulfate weathering. Soil Science Society of America, Special Publ. No. 10, Madison, WI.

Mains, C., B. Williams, and W. Chain-Wen. 1999. Deckers Creek Stream Quality Inventory: Summary Report. Friends of Deckers Creek. Morgantown, WV.

Moreland, J. 1985. Morgantown District Industrial and Business Survey (1920). West Virginia Collection, West Virginia University, Morgantown, WV.

Natural Resources Conservation Service (NRCS). 2000. Supplemental Watershed Plan No. 1 and Environmental Assessment for the Upper Deckers Creek Watershed: Preston and Monongalia Counties, West Virginia. Report by NRCS, Morgantown, WV.

Neville, C.M., and P.G.C. Campbell. 1988. Possible mechanisms of aluminum toxicity in a dilute, acidic environment to fingerlings and older life stages of salmonids. Water Air Soil Pollut. 42:311-327.

Paul, J.W. 1904. Bureau of Mines: Annual Report for the year ending 1903. West Virginia Collection. West Virginia University, Morgantown, WV.

Rosseland, B.O., I.A. Blaker, A. Bulgar, F. Kroglund, A. Kvellstad, E. Lydersen. D.H. Oughton, B. Salbu, M. Staurnes, and R. Vogt. 1992. The mixing zone between limed and acidic river waters: complex aluminum chemistry and extreme toxicity for salmonids. Environ. Pollut. 78:3-8.

Skousen, J., C. Mains, and R. Hamilton. 2000a. You can’t judge a stream by its color. Handbook for stream characterization with accompanying video. West Virginia University Extension Service, Morgantown, WV.

Skousen, J., A. Sexstone, and P. Ziemkiewicz. 2000b. Acid mine drainage control and treatment. Chapter 6. In: R. Barnhisel, R. Darmody, and L. Daniels (eds.), Reclamation of Drastically Disturbed Lands. 2^nd Ed. Agronomy Monograph 41, American Society for Agronomy, Madison, WI.

Sparling, D.W., and T.P. Rowe. 1996. Environmental hazards of aluminum. Reviews of Environ. Contamination and Toxicology 145:1-127.

Strausbaugh, P., and E. Core. 1977. Flora of West Virginia. 2^nd Ed. Seneca Books, Inc. Grantsville, WV.

Teti, M.J. 1974. Surface water geochemistry of the Deckers Creek drainage basin. M.S. Thesis. Department of Geology and Geography, West Virginia University, Morgantown, WV.

Thelin, R., and G. F. Gifford. 1983. Fecal coliform release patterns from fecal material of cattle. J. Environ. Qual. 12:57-63

Wiggins, B.A. 1996. Discriminant analysis of antibiotic resistance patterns in fecal streptococci, a method to differentiate human and animal sources of fecal pollution in natural waters. Appl. Environ. Microbiol. 62:3997-4002.

Wood, S.C., P.L. Younger, and N.S. Robins. 1999. Long-term changes in the quality of polluted minewater discharges from abandoned underground coal workings in Scotland. Quart. J. of Engineering Geology 32:69-79.

Young, K.D., and E.L. Thackston. 1999. Housing density and bacterial loading in urban streams. J. Environ. Engineering 125(12):1177-1180.

Younger, P.L. 1998. Coalfield abandonment: geochemical processes and hydrochemical products. p. 1-29. In: Energy and the Environ. Geochemistry of fossil, nuclear, and renewable resources. MacGregor Science, University of Newcastle, UK.

Picture 1. Headwaters of Deckers Creek.

Picture 2. Kanes Creek inflow to Deckers Creek.

Picture 3. Dillan Creek inflow to Deckers Creek.

Picture 4. Masontown site and location of major sewage input.

Picture 5. Greer site is below the Greer limestone quarry.

Picture 6. Dellslow site.

Picture 7. Richard Mine discharges very poor water into Deckers Creek.

Picture 8. Tramps site is downstream from Richard Mine discharge.

Picture 9.Morgantown site is under the High Street Bridge.