Douglas Abandoned Mine Land Project

Jeff Skousen

WVU Professor of Soil Science

Land Reclamation Specialist

 

History

"Coketon" is the term for the central mining facility of the Davis Coal & Coke Company between Thomas and Douglas, WestVirginia. Located along the North Fork of the Blackwater River in Tucker County, Coketon was an integral part of the vast and productive industrial complex of Henry G. Davis. In 1884, the railroad reached Thomas where Davis Coal & Coke had already begun active mining operations. Around the turn of the century, approximately 400 houses centered around the roundhouse and machine shops, and the valley was home to about 1,500 people. The community of Thomas boasted an opera house, hotels, banks, schools, and fraternal orders uniting 18 nationalities into the little city.

In the center of the Davis operation was a large coking facility on a mile-and-a-half stretch of the North Fork between Thomas and Douglas. An 1887 experiment with two ovens convinced the coal company that a coking facility would be very profitable. The coal company converted raw coal into coke, the purest form of carbon and the most important by-product of coal. Coke was the premier "reducing agent" fuel in the world at the time, capable of smelting iron ore rapidly into steel through the Bessemer process. In the late l800's, coke was produced by baking coal in huge stone or brick ovens until its impurities were driven off. The types of impurities driven out of the coal were a function of the amount of oxygen allowed into the oven (which was controlled by doors on the front of the oven) and the resultant temperature in the oven. It required nearly two tons of raw coal to produce one ton of coke. Long rows of "beehive" coke ovens, linked by tracks to the mine and tipple, burned night and day, tended by hundreds of laborers.

Eventually, Coketon contained 600 ovens. The Davis Coal & Coke Company's cokeyards employed about 150 men and burned for 250 days a year. By 1900, Coketon made Tucker County the third largest coke-producing county in the state. In 1904 alone, Coketon produced 200,000 tons of coke. During each year from 1915 to 1921, the 15 mines near Coketon shipped over 1 million tons of coal, making it the sixth most productive operation in West Virginia.

In 1915, a change in mining technology revolutionized the steel-making process, thereby eliminating the need for coke ovens at the mine site. By 1919, there was no coke production whatsoever in Tucker County, leaving the long banks of obsolete coke ovens unused. Coal, however, was still mined at the site in record quantities. From 1920 through the 1940's, the company continued production. As the seams were worked out and the mines closed, the population slowly declined, and the facility slowly began to shut down. By 1950, only two mines, #36 and #40, were still working and tonnage had fallen to 100,000 by 1954. By 1956, underground mining had ceased altogether with a few surface mining operations producing coal through 1965.

Although time and vandalism have eroded the Coketon complex, significant ruins are nevertheless extant. The town of Thomas retains much of its architectural integrity, featuring the company store and office building, and numerous well-kept and relatively unaltered miners' houses. Poured and cut stone and masonry foundations remain from the power house, as do the ventilation fan housing and tipple support pillars. Railroad trestles and graded railbeds line the North Fork of the Blackwater. Several mine portals stand open, including the #29 portal. The Chief Inspector of the West Virginia Department of Mines observed in his 1904 Annual Report that "an unusual amount of water is generated at #29, and the drainage is not very good."

The most significant and striking cultural resources of the site are the rows of coke ovens which line both sides of the valley. An entire bank of ovens stands free in the middle of the site (Picture 1), while both walls of the hollow are lined with the brick and stone ovens. The roadbed of the Western Maryland Railroad extends some miles to Hendricks, and features fine cut stone bridges and gorgeous vistas of the canyon below. (The preceding information was excerpted from Stuart McGehee, 1992, "Coketon: Documentation of Historic Resources")

Land Reclamation Construction Highlights

The Douglas Highwall AML Project contained approximately 62 acres of steeply sloping mine spoil and refuse, and roughly 4,200 linear feet of highwall that was vertical from 35 to 55 feet in height (Picture 2). Open and collapsed mine entries were also within the project limits. Reclamation of the site included these specifics (Pictures 3 and 4):

• 4,200 feet of highwall were eliminated;
• 360,000 cubic yards of spoil and refuse material was excavated, handled, and regraded;
• 62 acres of land were backfilled and revegetated;
• 5,300 feet of V-shaped subdrain were constructed on the pavement of the surface mine bench at the base of the highwall and covered during backfilling;
• 555 linear feet of underdrain were installed in various locations throughout the job,
• 550 linear feet of riprap were placed for erosion control along the regraded slope along the river;
• 1,500 linear feet of trapezoidal-type, fabriform-lined ditch were installed to receive discharges from the subdrains prior to outletting into the river;
• 7,600 linear feet of silt control were placed on the project,
• 6 wet seals and/or modified seals were installed on open mine portals; and
• the county road above the #29 mine entry was stabilized using 150 tons of stone and 53 cubic yards of grout.

Possible Treatment Systems for the Acid Mine Drainage

The #29 Mine Portal emits acid mine drainage from old deep mine workings of the Davis Coal & Coke underground operations. Depending on the season of the year, acid mine drainage flows out of the mine at the rate of 300 to 5,000 gpm. Analysis of the water has given the following range of water quality:

flow	200 - ???? gpm
pH	2.8 - 3.7
Acidity	400 - 600 mg/L
Total Iron	25 - 40 mg/L
Ferric Iron	11 - 17 mg/L
Ferrous Iron	14 - 23 mg/L
Manganese	5 - 11 mg/L
Aluminum	30 - 55 mg/L
Sulfates	700 - 800 mg/L
Dissolved Oxygen	<1 - 4 mg/L

Treating this water at an average flow of 500 gpm and 500 mg/L acidity with any one of the conventional AMD chemical treatment systems would cost between $60,000 and $200,000 per year.

Wetlands have been used to treat AMD because metals in acid mine drainage can be physically filtered by adsorption to organic materials, and the metals can also be oxidized/reduced by microbial reactions in the wetland and removed by precipitation. These processes remove metals from the water and neutralize some of the acidity.

Anoxic Limestone Drains (ALD) are used to treat AMD that has low ferric iron concentrations (less than 5 mg/L), and also low dissolved oxygen levels (less than 1 mg/L). Due to the variable ferric iron concentrations and moderate oxygen levels in the Douglas portal water, an ALD was unsuited to treat this drainage by itself.

Alkalinity-Producing Systems (APS) combine the use of an ALD and anaerobic compost wetlands. Under current designs, ponded water about 3 to 6 feet in depth overlies an 18-inch layer of compost, which is over an 18- to 24-inch layer of limestone. Acid water is ponded over the materials and the head created by the column of water forces the water through the organic material to filter out or precipitate ferric iron and to consume oxygen through organic matter decomposition. Alkalinity may be generated through microbial sulfate and iron reduction. The acid water, now low in dissolved oxygen and ferric iron after passing through the organic substrate, is then directed down into the layer of limestone under the organic matter or through pipes into a conventional ALD.

The Treatment System at the Douglas AML Site

During the planning and engineering phases of this AML project, the decision was made to treat this water with a combination of passive systems. The system designed at the Douglas Highwall project is different from other passive treatment systems. The system has two phases and employs an aerobic/anaerobic wetland at the surface, underlain by varying amounts of limestone (Figure 1). This system is different because the amount and thickness of organic matter is much greater and the water depth is not so deep at the surface, so wetland plants can be established in the organic substrate. Such a system with this design has not been constructed on an actual mine site. The system is designed to facilitate four processes:

1. Remove ferric iron in the water by: a) adsorption to or precipitation in the organic substrate in the surface of the treatment system, and b) microbial ferric iron reduction producing soluble ferrous iron in the anaerobic sediment;
2. Remove the dissolved oxygen in the water by consumption of the oxygen through organic matter decomposition;
3. Add alkalinity to the water by: a) microbial sulfate reduction in the anaerobic zone of the organic material, and b) limestone dissolution from the ALD, and
4. Precipitate metals in the catchment basin before the water is discharged into the North Fork of the Blackwater River.

Water Treatment Construction Specifications

The AMD treatment system was constructed by digging a large trench along the old railroad grade. The first cell of the system was 1,225 feet long by 8 feet wide and 6 feet deep (1,225 x 8 x 6 = 58,800 cubic feet or 2,178 cubic yards). About 2 feet of gravel-sized limestone (#57 size and >90% calcium carbonate equivalent) were placed on the bottom with 4 feet of organic material (a peat, hay, and soil mixture, 50:40:10) over the top. Cell I was filled with about 880 tons of limestone and about 1,450 cubic yards of organic material.

Cell II was 1,375 feet long by 30 feet wide and 8 feet deep (1,375 x 30 x 8 = 330,000 cubic feet or 12,220 cubic yards). About 6 feet of 2- to 4-inch limestone was placed on the bottom with 2 feet of organic material over the top. This cell was filled with about l2,000 tons of limestone and about 1,000 cubic yards of organic material (Pictures 5 and 6). The organic substrate was planted with cattails and other wetland vegetation (Pictures 7 and 8).

Operation and Function

This AMD treatment system will function by introducing approximately 240 gpm of acidic water from the #29 Portal into the ALD/Wetland System. Much of the ferric iron in the water will precipitate or be adsorbed onto the organic material in the surface of the wetland. If ferric iron comes into contact with limestone, ferric hydroxide will form around the limestone and make it less reactive for neutralization. Another portion of the ferric iron may be reduced by microorganisms forming soluble ferrous iron. Ferrous iron does not form a coating on limestone at pH 7 or less. It is important that the ferric iron be removed or reduced in the water as it moves downward through the organic substrate.

Another concern is the potential coating of limestone or plugging of limestone pores by aluminum hydroxides. Aluminum adsorption and/or precipitation should also occur in the organic matter substrate as the water pH reaches 5 or above. The small amounts of oxygen in the acid water should be scavenged by microorganisms decomposing the organic matter as the water moves through the organic substrate. With little or no ferric iron and little oxygen, the limestone in the ALD/Wetland System should dissolve and add alkalinity to the water.

At the end of the passive treatment system, a small aeration and metal precipitation pond was constructed to catch/trap the metals before the water is discharged into the North Fork of the Blackwater. The metal hydroxide sludge that accumulates in the pond will have to be periodically cleaned out.

Calculated Longevity

Based on 240 gpm and 500 mg/L acidity, the treatment longevity of this drain with its 13,000 tons of limestone (90% CCE and 75% dissolution) is estimated to be 50 years. Organic material has a finite capacity to adsorb metals, so the longevity based on the capacity of the organic material may be less than 50 years. Wetland plants have been transplanted into the organic material and these plants will deposit additional organic material annually through die-back. Little information exists on the longevity of organic substrates in AMD-treating wetlands.

Sampling and Analysis of Water

The Douglas WALD was sampled monthly from June till November 1994, quarterly until September 1995, then at periodic intervals since then. Water samples were taken by allowing water to flow into a container prior to pumping through a YSI Model 3500 sample chamber (Yellow Springs Instruments, Yellow Springs, OH) equipped with Eh, pH, EC and temperature probes. Water samples for metal analyses were filtered (0.45um filters), acidified with HCl (0.24M final concentration), and stored in sealed vials at 4^oC until analyzed. Total Fe, Mn, Al, Ca and Mg were analyzed using a Perkin Elmer Plasma 400 Inductively Coupled Plasma Spectrometer (ICP). Ferrous iron was quantified using a modification of the ferrozine procedure (Stookey 1970). Standards were prepared using FeSO₄ ^. 7H₂O prepared in 0.25M HCl. Ferric iron was calculated by difference as Fe (total) - Fe²⁺. Sulfate was measured using a Varian model 2510 HPLC pump equipped with a Vydac 3021C 4.6 anion column, an Alltech 28069 anion guard column, and a Milton Roy LDC Conductometer. Total acidity and alkalinity were determined by fixed endpoint titration to pH values of 8.3 and 4.2, respectively, using a Titralab VIT-90 automatic titrator (Radiometer-America, Westlake, OH). Water quality data have also been collected upstream and downstream of the reclamation activities in the area.

Results of the ALD/Wetland System at Douglas

The entire project cost at Douglas including all reclamation activities and water treatment systems was $1.4 million. The wetland-drain system costs about $400,000.

The system raised pH of the 240 gpm flow from an average of 3.0 to a maximum of 7.3 during the first year (Table 1). After one year, effluent pH declined from 5.2 after 13 months (9/95) to 3.0 after 19 months (3/96). Effluent pH after approximately four years of operation has been about 3.5. Acidity decreased from 300 mg/L as CaCO₃ to an average net alkalinity of 127 mg/L as CaCO₃ during the first year. During the last three years, acidity values of water exiting the system averaged 169 mg/L as CaCO₃. Even though alkalinity is not measurable in the water, the system has caused a reduction in net acidity.

Iron and Al, the metals of primary concern, did not exit the wetland-drain system during the first year and ferrous iron was not generated in the system by iron-reduction reactions. Manganese was also removed by the system during the first year. It is evident that oxidizing conditions prevailed as the pH and mV values measured in the wetland were insufficient for Mn precipitation. It is estimated that about 13 tons of Fe and about 18 tons of Al were retained in the wetland-drain during the first year. Since that time, only about half that amount of metals have been accumulating (about 6 tons of Fe and 10 tons of Al per year). Up to this time, it appears that these precipitated metals are remaining in the wetland and no movement out of the wetland is evident.

The high mV and dissolved oxygen readings in water throughout the system confirmed our observations of water flow in the system (data not shown). Most of the water flowed across the surface of the wetland in Cell I, rather than migrating downward through the organic matter into limestone. The low permeability of the organic substrate lead to short circuiting of the flow and little contact with organic matter and limestone was attained with the majority of the water. These data and our observations suggest that the Douglas system acted as an aerobic, Fe-oxidizing system rather than an anaerobic, Fe-reducing system, as originally intended.

After a few months of operation (in November 1994), we attempted to reduce the surface flow and encourage ponding and downward movement of the water into the organic substrate by installing a series of hay bale dikes every 10 m in Cell I (Picture 9). However, without an outlet for the water below the limestone and organic substrate (such as a drainage system), the water simply ponded behind the dikes and exited at the dike's lowest point. Beavers have since taken up the task of ponding the water in the system (Picture 10).

Although the effluent water quality was dramatically improved during the first year, the longevity of the system was compromised since Fe and Al in the water precipitated in the system. The system has not generated measurable alkalinity, but metals are continuing to precipitate in the system, thereby reducing the acidity of the water and the metal load to the river.

Assuming the average influent acidity value was 300 mg/L as CaCO₃ and the average effluent acidity value was 170 mg/L as CaCO₃ at the 240-gpm flow, an acid load reduction can be estimated. Approximately 158 tons of acid per year entered the river before treatment, while 89 tons of acid per year enter the river after passing through the wetland-drain system. This represents 69 tons less acid per year (or about a 44% reduction), and multiplying that amount for the past three years equals more than 200 tons of acid that have not entered the river. The system may continue to reduce the acidity of the water to this level for many more years. So while the system may not be introducing net alkaline water into the river, significant amounts of acid and metals are not entering the river due to the system (Skousen et al., 1999).

When comparing the acid removal of this wetland-drain system to other successive alkalinity-producing systems (SAPS), we see an interesting trend. The alkaline-producing systems (SAPS) are predicted to remove about 15 to 20 g/m²/day of acid and this value represents the commonly accepted target that most wetland builders expect. When calculating the acid removal per square area per day for the Douglas system, we find that this system is removing 34 g/m²/day of acid. Robert Nairn, a wetland designer, stated that this is the highest removal rate for any alkaline-producing system (SAPS) or wetland he has seen.

The cost to treat the acid mine drainage at Douglas with this wetland system is estimated to be $828 per ton of acid neutralized. This cost per ton is almost equivalent to NaOH chemical treatment cost, which is estimated at $856 per ton. While this site had a high per ton cost, this wetland-drain system also treats a higher flow than many passive treatment systems. The installation cost of $400,000 was also higher than originally estimated because of modifications that occurred. With continued removal of metals and acidity from the water, the cost of treatment will continue to go down.

Acknowledgments

Alan Sexstone and Keith Garbutt, both professors at West Virginia University, were co-investigators on this project. I also acknowledge the extensive work of Joe Calabrese, John Cliff, and Pat Sterner in water collection and analysis. Thanks also are extended to Paul Ziemkiewicz, John Sencindiver, and DK Bhumbla for their help. I appreciate the West Virginia Division of Environmental Protection and Triad Engineering, Inc., for initiating West Virginia University's involvement in this project. Special appreciation is extended to former Division Director David Callaghan, Dave Broschart, Marshall Leo, and David L. Smith. Breckenridge Corporation was the contractor, and thanks go to Alan Shreve and Tim Stump for their help and assistance. I also thank Sheila Vukovich and Mike Sheehan for water-quality data for the Blackwater River and to Buffalo Coal Company for aerial photographs. Funding for conducting the water analysis of the wetland system was from the U.S. Bureau of Mines and the National Mine Land Reclamation Center at West Virginia University.

Table 1. Water quality of the North Fork of the Blackwater River upstream and downstream of the Douglas and Albert Abandoned Mine Land Projects (including the wetland-drain system at Douglas), and the water quality of the acid mine drainage going into and exiting the wetland-drain over time.
_____________________________________________________________________________________

Table 1 continued. North Fork of the Blackwater River Data.
_______________________________________________________________________________________________

______________________________________________________________________________________