| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a 629 East Rolling Ridge Drive, Penn Eagle Industrial Park, Bellefonte, PA 16823
b Plant and Soil Science, West Virginia University, Morgantown, WV 26506
* Corresponding author (jskousen{at}wvu.edu).
Received for publication April 8, 2003.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: AMD, acid mine drainage
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Acid mine drainage (AMD) is a serious problem in areas of extensive surface and underground coal mining, such as the Appalachian region, where pyrite and other metal sulfides are found within the coal and associated rocks. About 10000 km of streams have been affected by AMD in Pennsylvania, Maryland, Ohio, and West Virginia (USEPA, 1995). Many mines currently discharging AMD were operated and abandoned before enactment of the Surface Mining Control and Reclamation Act (SMCRA) of 1977. The act provided standards for environmental protection during mining operations and placed the responsibility of AMD control and treatment on the operator (United States Government, 1977). The SMCRA also provided a means for reclaiming abandoned mines by taxing current coal operators, which generates funds for abandoned mine land reclamation programs. Even with millions of dollars spent in reclaiming abandoned mine lands, these abandoned mines still generate more than 90% of the AMD in streams and rivers in the region and most of this acidic drainage flows from underground mines (Faulkner, 1997; Zipper, 2000).
Because these sites were abandoned before 1977, no company or individual is responsible to treat the water, and therefore the receiving streams are continuously polluted and severely affect the aquatic ecosystem. High flows and high levels of pollution (high acidity and metal concentrations) require the use of chemicals for treatment, which tend to be expensive and labor-intensive (Skousen et al., 2000). Costs for chemicals, dispensing equipment, electricity for pumps, and manpower all add up to significant public expense if the treatment entity is a government agency or utility. Perhaps the largest cost to the public is the unavailability of the water resource for use and its accompanying impaired aesthetics and degradation. Therefore, simple and inexpensive treatment approaches (such as passive treatment systems; Skousen et al., 1999) are being sought as well as a better understanding of the natural processes within mines that affect water quality over time.
An understanding of the behavior of acid-producing materials within abandoned mines would allow an estimate of the longevity of the acid discharge, which will aid in determining remediation strategies and the short and long-term costs of treatment. However, the changes in flow and water quality over time from surface and underground mines are not well documented.
Surface mining generally removes 90% or more of the coal (which often contains the highest sulfide content and hence the acid-producing potential) thereby leaving little in the backfill for continued reaction and acid generation. Pyrite is often dispersed among other rocks above and below the coal seam. The pyrite-bearing rocks and coal left behind are broken apart by blasting giving high surface area, and the acid products are leached fairly rapidly, typically within 10 to 20 yr (Meek, 1996). Carbonate rocks within the overburden may neutralize some or all of the AMD generated during surface mining. Special handling of toxic materials may reduce the amount of pyrite oxidized, and the addition of alkaline material during mining may neutralize acid in situ, both of which decrease the total acid load coming from the site (Brady et al., 1990; Perry and Brady, 1995; Rich and Hutchinson, 1990; Rose et al., 1995; Skousen and Larew, 1994). During the ensuing 20 yr after reclamation, discharge water quality may reach pre-mining levels. Overburden analytical techniques such as acid–base accounting help predict the overall potential for AMD during surface mining and reclamation, and can help regulators determine when surface mining should or should not be allowed (Skousen et al., 2002).
Acid discharge from underground mines usually lasts much longer than from surface mines (Wood et al., 1999). For underground mine discharges, a distinction is made between those that become inundated or flooded with water (the coal seam is located below the regional water table and the mine is called "below-drainage") and those that do not fill with water because they are situated above creeks and rivers of the area (the coal seam is located above the regional water table and the mine is called "above-drainage") (Fig. 1) . This distinction is critical because acid-producing materials composed of pyrite react at much slower rates and produce only small amounts of AMD when left in an anaerobic condition underground (Evangelou, 1995; Fennemore et al., 1998). Therefore, flooded below-drainage underground mines tend to have a finite life for discharging AMD because oxygen depletion limits acid generation.
|
Other researchers have observed that the most severe drainage occurs within the first few decades after closure and even the largest underground mine systems settle to lower pollution levels within 40 yr. For mines in the UK, a neutral pH was reached within 30 yr, and after 40 yr the pH remained neutral and the iron concentrations decreased from 200 to <40 mg L–1 (Wood et al., 1999). Jones et al. (1994) also showed that water from flooded underground mines in Pennsylvania changed from acidic to neutral over a period of decades.
Lambert and Dzombak (2000) located three underground discharges in the Uniontown Syncline of Pennsylvania with distinct flooding histories. This study used underground mines in the Pittsburgh coal seam, one of the coal seams we sampled in our study. Water quality measurements had been taken in 1974 and 1999 in each mine. A flooded below-drainage mine closed in 1934 (40 and 65 yr had passed since closure) had a pH of 6.0 in 1974 and 6.4 in 1999, Fe decreased from 45 to 25 mg L–1, and sulfate decreased from 1700 to 1000 mg L–1 (net alkaline water). In a flooded below-drainage mine that was closed in 1970 (instead of 1934), pH increased from 3.1 in 1974 to 5.9 in 1999, Fe decreased from 140 to 70 mg L–1, while sulfate decreased from 2000 to 900 mg L–1 (changed from strongly acidic water to slightly acidic water). Therefore, the researchers concluded that underground mine water quality changed from acidic to alkaline within 30 yr after closure and flooding in their geologic setting. Water pH from an unflooded above-drainage mine closed in 1934 was 3.0 in 1974 and 3.5 in 1999, while Fe decreased from 10 mg L–1 in 1974 to <2 in 1999 and sulfate declined from 800 mg L–1 in 1974 to 600 mg L–1 in 1999. Water in all cases was net acidic from the unflooded mine. Therefore, unflooded above-drainage areas improved in drainage quality, but still remained net acidic (albeit with low metal concentrations) 60 yr after closure. Other researchers have found similar results in this region (Brady et al., 1998; Capo et al., 2001).
Donovan et al. (2000) monitored the water quality of a large underground mine pool in a setting similar as that of Lambert and Dzombak (2000). The Montour mine, a Pittsburgh coal seam underground mine in Pennsylvania, had a section flooded in 1970 and also a section that was flooded in 1982. Water conditions in the connected mine pool were strongly acidic (pH of approximately 3.0, acidity of 44 mmol H+ L–1) for the first two years after the final flooding in 1982, after which acidity began to decline exponentially. Surprisingly, seven years after flooding (five years after peak acidity concentrations), the water in the mine pool became net alkaline and has stabilized at a pH of 6.4, net alkalinity of 4 mmol OH– L–1, and iron around 60 mg L–1. The good quality water in the 1970 flooded mine probably caused rapid neutralization of the water in the 1982 flooded mine.
Younger (1997) divided the acid load flowing from underground mines into two categories. "Vestigial" acidity is associated with the first-time flushing of acid products from the mine during initial abandonment and flooding. "Juvenile" acidity is produced from ongoing pyrite oxidation due to fluctuations in the water table. Juvenile acidity, he states, can persist for hundreds of years depending on the pyrite content and hydrology of the underground mine system. The longevity of AMD at a given site is dependent on the rate of depletion of both the vestigial and juvenile acidity. Younger (2000) found that flooded underground mines in the UK stabilized at pH 7 with gradually decreasing iron concentrations.
In above-drainage underground mines, the mine voids and acid-generating materials are continually exposed to seasonal high and low water levels in the mine or mostly unflooded conditions throughout the year, and tend to be recharged rapidly by infiltrating rain and snowmelt. Furthermore, the water moving into the mine voids can be discharged just as rapidly out of the mine depending on its connectivity with surface portals or seeps. The mine atmosphere contains sufficient oxygen to sustain pyrite oxidation because the mine atmosphere interacts with the outside atmosphere via fractures and unsealed openings, especially in response to changes in atmospheric pressure (i.e., they "breathe"). Therefore, oxygen and water are not limiting factors for continuing pyrite oxidation and transport of reaction products. Indeed, these quick recharge–discharge and breathing conditions are ideal for acid generation and pyrite reaction continues until the pyrite supply either becomes fouled by acid product coating (armoring) or becomes exhausted. Under these optimized oxidizing and flushing situations, it is possible that discharges could be contaminated for decades or centuries depending on the pyrite supply and flushing events.
In above-drainage mines, water usually discharges out at the down-dip side of the mine at the portal or at other low elevation areas in the mine through thin coal barriers or fractures. Regardless of the discharge point, the mine water continually flows out without fully flooding the mine and acid-generation may continue for decades until the pyrite is either exhausted or coated, which limits surface area for reaction (Lambert and Dzombak, 2000; Younger et al., 1997). In these situations, the rate of water quality improvement over time (or acidity decline) is highest where the recharge rate of the mine is high, water flow out of the mine is high, residence time is low, and the mine volume is small (Younger, 1997). The fluctuating water level and pooling effect due to seasonal variations in precipitation, however, aid acid generation. During low water levels, pyrite oxidation forms iron hydroxysulfate solids, which precipitate on coal and rock surfaces due to evaporation. Hydrated ferrous and ferric sulfate minerals such as melanterite, rozenite, copiapite, and jarosite have all been identified as precipitated minerals in coal mines (Nordstrom, 1982; Younger, 2000), and all of these minerals provide short-term storage of metals, acidity, and sulfate that are soluble. When the water level rises, these acid products are dissolved, released into the mine pool, and flushed out of the mine during seasonal recharge times. Pyrite oxidation can continue to occur on the wet, oxidized mineral surfaces, producing an optimized and continuing cycle of acid generation (Younger, 1997). Therefore, these above-drainage underground mines generally will discharge poor water quality for much longer periods than flooded mines because more of the pyrite is continuously exposed to a mixture of air and water to form acid products, and released.
Much discussion has occurred on the seasonal changes in water quality from above-drainage underground mines, including changes in pH, acidity, and metal concentrations. Pigati and Lopez (1999) reported that the observed seasonal variations in water chemistry from the above-drainage 126-ha Majestic Mine in Ohio supported the "spring flush" theory (Johnson and Thornton, 1987). Acidity and metals generated and deposited on surfaces during summer and fall (seasonal dry months in this region) were largely flushed with late winter–early spring snowmelt and rainfall. Therefore, during the high recharge and discharge season of February to May, the high water flows carried high concentrations of acidity and metals. This is opposed to the "high concentration at low flow" hypothesis, where the worst drainage from underground mines happens at base flow conditions (dry months) where acid and metals are thought to be concentrated and not diluted by surface water during recharge events (Skousen and Ziemkiewicz, 1996).
The Omega Mine, a 68-ha above-drainage underground mine in the 1.4-m-thick Upper Freeport coal seam in northern West Virginia, was mined during the 1980s. In 1992, breakout of AMD caused mine closure, and flows and acidity concentrations were recorded continuously for the ensuing six years (GAI Consultants, 2001). From 1993–1999, average year-round water flow was 1.4 L s–1 (22.6 gallons min–1) and acidity was 97 mmol H+ L–1. Water flow from the mine averaged 2.7 L s–1 (43 gallons min–1) during February to May (1993–1998), and only 1.1 L s–1 (18 gallons min–1) during June to January (1993–1998). Acidity was highest during February to May (approximately 90 mmol H+ L–1) compared with about 76 mmol H+ L–1 during June to January. Above-drainage mines, based on these and other examples, appear to discharge their worst acid and metal concentrations during high flow seasons (spring flush), with improved water quality during dry periods.
It is clear that discharge chemistry from underground mines is affected by several factors, and efforts to determine the changes in water quality over time require accounting for these factors. First, the season of sampling is important to consider, with two distinct periods being documented in our region: February to May and June to September. Flows in October, November, December, and January from underground mines are especially hard to predict because these months alternate between very high flows like those of February to May and very low flows like those from June to September (GAI Consultants, 2001; Pigati and Lopez, 1999).
Another important factor in mine water chemistry is the coal seam, and more specifically the pyrite content of the mined coal (Younger, 2000). Each coal seam is unique with relatively predictable chemical and physical features, which affects the magnitude of the discharge water quality.
The mining method and degree of coal removal within a mine are other variables affecting discharge chemistry. Room and pillar underground mining (the most common method in this area) left about 50% of the coal as support for the roof (Reece et al., 1978). After abandonment, this coal continues to weather and crack from the pillar and oftentimes the pillar completely collapses, allowing more pyrite in the coal pillar and overlying rocks that have dropped into the void to react. In "retreat" mining, many of the coal pillars left to support the roof are removed as the mining equipment is withdrawn, allowing more roof rock to subside into the mine voids. The coal pavement or floor rock also can contribute pyrite surfaces for reaction, depending on its flooded condition and slope. Remining underground mines by surface mining has the potential to improve preexisting acid discharges by removing coal pillars and then reclaiming the site to current reclamation standards, which often includes mitigating any AMD potential (Hawkins, 1994; Richardson and Doughterty, 1976).
Acid discharges from abandoned mine sites can be affected by subsequent, adjacent surface mining (Skousen et al., 1997) or other nearby surface disturbances. The flow may decrease because the surface overlying the recharge area has been reclaimed and vegetated, thereby decreasing infiltration into the underground mine, or surface fractures or subsidence holes may have been repaired and closed off. Any of these surface disturbances could effectively decrease the flow rate into and therefore out of the mine. Adjacent surface mining and blasting may also cause collapse of the roof in portions of the mine and reduce the void space, thereby changing flow paths or altering interconnection of certain areas. The degree of disturbance in a mine is nearly impossible to measure (down hole cameras have been used to look into underground mines, but they allow only a small peek into the void space) and only observable surface disturbances overlying or adjacent to the underground mine provide evidence that such disturbance effects exist. Even so, the ensuing effect of surface disturbances on mine water chemistry is difficult to predict over time.
A research project was initiated in 1968 which identified and sampled more than 150 underground mine discharges in the northern West Virginia coal region. Most of these discharges were coming from unflooded above-drainage underground mines in the Upper Freeport and Pittsburgh coal seams. We revisited many of those sites in 1999 and 2000 and collected and analyzed the water from about 75 of the same discharge points. Based on the completeness of the 1968 data, 44 were chosen for further analysis. Twenty of these 44 sites had been sampled and the water analyzed in 1980, which provided an intermediate sampling time to assess water acidity and iron concentrations from the mine. From this data set, our objective was to determine the changes in water quality during this 30-yr period.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The Pittsburgh coal seam is the lowest stratum of the Monongahela Group in the Pennsylvanian System. The seam has 1.5 to 2% sulfur and an ash content of 6%. The Pittsburgh coal is composed of alternate layers of coal and black shale. A typical Pittsburgh coal cross-section shows a 1-m layer of pure coal, a 0.7-m layer of bone coal or slate, and another 2-m layer of good-quality coal. The Pittsburgh coal along the Monongahela and Cheat Rivers is located close to the surface, and can be mined by surface mining methods or shallow underground mines (Hennen and Reger, 1914). In this region, few overlying limestone materials are available within 30 m above the coal seam to neutralize the high amounts of acid-producing material in this coal and associated rocks.
The Upper Freeport coal seam is the topmost stratum of the Allegheny Formation in the Pennsylvanian System. Upper Freeport coal contains <1.5% sulfur, and an ash content from 8 to 12%. It is a multiple-bedded seam that is divided into a top coal and bottom coal, separated by a shale interlayer, all averaging a total of 2 m in thickness (Hennen and Reger, 1914). The strata above the Upper Freeport coal contain several massive sandstones and some shales. Limestone or alkaline-bearing rock units are not generally found within 50 m above the Upper Freeport coal in this area, so very little overlying geologic material is available for acid neutralization (Hennen and Reger, 1914).
1968 Study
A previous research project was conducted during the summers of 1968–1970 (June–September) where field crews were sent out to identify all coal mines within the Monongahela River basin and to sample their discharges. Each crew worked from 7.5-min U.S. Geological Survey topographic maps on which they outlined mine boundaries and indicated mine openings. Field sheets were also completed at each site with location and overburden information. Sites with a discharge were identified on the maps, flow rates were determined, and the water was sampled. The flow rate was measured when possible with a bucket and stopwatch. For larger flows, the crew installed a V-notch weir and measured flow rate. These values were recorded on the field sheet. In the field at the time of water collection, the pH of the discharge was measured using an electrometric pH meter, and temperature was checked with a lab-grade thermometer. These values were recorded on the field sheet.
Two water samples were taken at each discharge in this early study: (i) a plastic, 1-L bottle was filled with water, put on ice, and then analyzed in the laboratory for acidity, alkalinity, conductivity, sulfate, and pH; and (ii) a 50-mL glass bottle was filled, treated with acid, and then analyzed in the laboratory for metals (total iron, manganese, aluminum). Water samples were delivered to the laboratory each Friday where they were analyzed using methodology from the latest edition of Standard Methods (American Public Health Association, 1965). Water analyses were monitored for accuracy and precision by running periodic samples of reference standards (G. Bryant, personal communication, 1999).
1999–2000 Study
After obtaining the 1968 maps, we located the point discharges marked by the 1968 crew on the topographic maps during the summers of 1999 and 2000 (June–September) and correlated these to underground mine maps and areas (Table 1). Underground mine maps were obtained from the West Virginia Geologic and Economic Survey in Morgantown, West Virginia, underground mine boundaries were determined, and mine size was determined with a digital planimeter (Sokkia Corp., Overland Park, KS). The year in which each mine was opened was also determined from records of the West Virginia Office of Miners Health Safety and Training.
|
Discharges were sampled as close to the mine portal as possible. Flows were calculated using a measured cross-sectional area and flow velocity. Two water samples were taken at each sample point: (i) a 250-mL unfiltered sample was taken for general water chemistry (pH, conductivity, total acidity and alkalinity by titration, and sulfate); and (ii) a 25-mL filtered sample was acidified to pH of <2 with 0.5 mL concentrated nitric acid and used to determine metal concentrations.
Water pH, alkalinity, and acidity were determined by a Metrohm pH Stat Titrino System (Brinkman Instruments, Wesbury, NY). Conductivity was measured using a Model 115 conductivity meter (Thermo Orion, Beverly, MA). The metal analysis was preformed using a Plasma 400 inductively coupled spectrophotometer (PerkinElmer, Wellesley, MA). Sulfate was measured turbidimetrically by flow injection analysis (Lachat Instruments, Milwaukee, WI).
The West Virginia Division of Water Resources also conducted periodic sampling and analyses of underground mine discharges in this area (West Virginia Division of Water Resources, 1985). We accessed their data and found that 20 of their sample sites matched our discharges sampled in 1968 and 1999–2000. Therefore, we used their water quality analyses as an intermediate data point between 1968 and 1999 to aid in estimating the rate of change (improvement) in water quality.
Statistical Analysis
A repeated measures analysis of variance was performed to test for changes in discharge water quality over time. We used a full model with the main effects of date, coal seam, disturbance, and all possible interactions as class variables using PROC GLM (SAS Institute, 1989). For the comparison between 1968 and 1999–2000, coal seam (Pittsburgh and Upper Freeport) and disturbance (yes or no) were the between-subjects variables. For comparisons between 1968, 1980, and 1999–2000, there were not enough degrees of freedom to make meaningful comparisons with disturbance as a model term. Therefore for the analysis of three dates, coal seam was the only between-subjects variable. The Greenhouse–Geiser criterion was used for parameters that did not meet the sphericity condition as determined using Mauchley's test (Huck and Cormier, 1996). Since our principal interest was the time dependent changes in discharge water quality, the assumption of homogeneity of variance was not tested in the between-subjects variables, and between-subjects comparison are neither provided, nor interpreted. For all statistical analyses, samples collected in 1999 were coded as if they were collected in 2000. Trends in discharge water quality improvements were plotted for the 1968, 1980, and 1999–2000 data.
To assess the potential effects of flow on discharge water chemistry and to determine the amount of variability within a sampling season, we gathered historical data from two discharges of above-drainage underground mines in this region with more frequent within-year sampling and examined seasonal changes in flow and concentrations. The Omega above-drainage underground mine is a 70-ha mine and is located in an adjacent watershed from most of our underground mines, but drains the same Upper Freeport coal seam as many of the discharges in this study. Six years of data (1993–1998) were available from this site (GAI Consultants, 2001). The T&T above-drainage underground mine drains an area of about 600 ha and is located in the same watershed and in the same Upper Freeport coal seam as most of the underground mine discharges used in this study. Seven years (1994–2000) of data were available from this discharge. Analysis of variance with year and month as categorical variables was used to test for a relationship between flow and discharge water quality using 12-mo versus 4-mo (June–September) data.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
) as this gives a basis for comparing changes in water quality over time. The % for each parameter was always larger than the RSD (Table 4). Percent differences more negative than the 4-mo % was used as an estimate of a conservative criterion for demonstrated improvement in water quality on our 44 discharge sites.
|
|
For all other parameters, there were significant improvements between 1968 and 1999–2000 (p = <0.001), with average improvements of 66% for total acidity and 79% for iron (Table 5). The date x coal seam interactions were due to significant differences between coal seams in 1968 and a lack of differences in 1999–2000, except for pH in 1999–2000 (Table 5). Average decreases in Pittsburgh coal seam iron (85%) and total acidity (79%) were larger than the decreases in Upper Freeport coal seam iron (74%) and total acidity (56%). This is probably due to the higher sulfur content of the Pittsburgh coal seam and possibly to the higher amount of shale associated with the seam compared with Upper Freeport coal. The shales weathered quickly releasing the acid products resulting in high initial acidity, iron, and sulfate concentrations that descended gradually over time. Upper Freeport coal had more sandstone, which reacted and released acid products more slowly than the shales.
Even though significant improvements in average iron and acidity concentrations were found, not every site showed improved water quality. Figure 3
is a cumulative frequency plot for the percent change in acidity and iron. All Pittsburgh sites showed some improvement (% change was <0), while one Upper Freeport site showed an increase in iron and five Upper Freeport sites showed increases in acidity. Using zero percent change (%) as the criterion for improvement is probably too liberal as it ignores the potential variability within a year that could bias the conclusions. Using a more conservative criterion for demonstrated improvement (–38.8 for acidity and –59.2 for iron; Table 2), 10 of 35 Upper Freeport sites did not improve in acidity and 11 of 35 sites failed to show improvement for iron. One Pittsburgh site also failed to show an improvement in iron.
|
|
|
|
Despite these overall trends, not all sites behaved similarly. When each site was plotted individually, all five Pittsburgh coal seam discharges showed some improvement in total acidity and iron (Fig. 4) . For acidity, all but one (Lake Lynn 1) showed an apparent curvilinear trend with time. For iron, only Cheat River 4 and Cheat River 5 showed this curvilinear trend (Fig. 4). For the 10 Upper Freeport discharges (out of 15) that showed improvements in acidity, most of those trends (8 out of 10) were apparently linear (Fig. 5) . For the Upper Freeport sites that did not show improvement in acidity, three of the five had an increasing linear trend with time. For the other two sites, acidity in 1980 was lower than in either 1968 or 1999–2000 (Fig. 5). For Upper Freeport sites with a decrease in iron concentration, only 4 of 11 showed an apparent curvilinear trend (Fig. 6) . In Upper Freeport sites that did not show improvement in iron, most were because the 1980 iron concentrations were the lowest.
|
We evaluated the effect of mine size on acidity and found no relationship in our discharge data (p = 0.38). While this result seems counterintuitive, there are some reasons that may account for this finding. Several of the discharges in our study were found to be coming from the same underground mine (based on old underground mine map locations). Examples of this are the Glade Run 4 and 5 discharges, and the Bull Run 1 and 2 discharges in Table 1. One would assume that discharges coming from the same underground mine (even though one discharge might be located around the hillside or in the next valley from the other) would be similar in quality. However, some of the discharges coming from the same mine were very different in quality, which resulted in poor correlations between mine size and acidity. Stratification has been noted from underground mine pools (Ladwig et al., 1984), but this stratification has been measured in flooded underground mines with long residence times. One certainly would not expect large differences in discharge quality from shallow, unflooded above-drainage underground mines, which are generally flushed within days of rain events. If large differences were found, then the residence time of the water in the underground mine was substantially increased, there may have been stagnant pools of water in the mine formed by roof caving or pillar collapse, the coal seam characteristics (pyrite or shale content) were drastically changed over a short distance to cause an increase in acidity concentrations, or the water drainage from the mine came from a different section of the mine. A hydrologic situation where residence times could increase may be with an undulating coal seam, where the coal seam dips and rises causing the creation of many small cells or pools of water. This effectively creates pockets of high acid water that may be discharged only during and after rainfall events. Other reasons for mine water variability within one mine are oxygen availability in distinct sections of the mine, which can vary dramatically with short distances, and the order of water contact with rocks in the mine (P.L. Younger, personal communication, 2003).
We found no relationship (p = 0.31) between the date the mine was opened and acidity. While a relationship would be expected to exist between time since mine closure (not mine opening date) and discharge quality, we were unable to accurately assess this relationship for a couple of reasons. The historical records were too sketchy to determine when all the mines ceased operation. We were able to find the opening date of all the mines because certificates and permits were required, but such paperwork was not necessary for closure. Therefore, even if a mine opened in 1945, the coal may have been mined for only a few years or for more than 30 yr. Mine size could help indicate the length of operation, but some mines worked slowly with few men and machines, while other mines operated quickly. So the closure date could have been 1950 or 1975, and one would predict the drainage quality to be better from a mine that was closed 50 yr ago compared with one closed 25 yr ago. Sometimes mines ceased operation during bad market or labor conditions, then reopened when conditions changed, or sections of the mine may have been closed off early in the life of the mine. All of these influences could confound the effect of mine opening or closure on discharge water quality.
While predictions for annual acidity declines or decay rates have not been made here, such as those made by Ziemkiewicz (1994), Younger (2000), and Demchak et al. (2001), these predictions have proved helpful in determining remediation strategies and costs. For example, given the scenario that the first 15 to 20 yr after closure will probably produce the highest acidity levels (as was found here), suitable chemical treatment systems can be installed to treat the low pH, metal-laden water. Intermediate levels of acidity are likely to be released during the next 10 to 20 yr, resulting in lower chemical costs and management. Accordingly, 30 to 40 yr after closure, acidity levels from these above-drainage underground mines tend to be much lower than initial concentrations and may be treated by low-cost passive systems.
Efforts must continue to determine water quality improvements from underground mines in other coal seams, and to better determine the effects of mine size and time since mine closure. The remediation strategies and cost predictions are based on our data as well as other research where water quality appears to improve within 40 yr after closure. Our data suggest that the water quality improvement predictions of Younger (1997) and Glover (1983) do not fit all the above-drainage underground mines we sampled, but certainly are appropriate for the flooded below-drainage mines they encountered in the UK and also those in the eastern USA.
There is still much work to be done on mine drainage improvement over time. It is critical to have a better understanding of the time needed for mine drainage improvement so as to plan suitable watershed restoration projects. Situations where the flow or the underground mine pool have been altered (due to sealing, grout injection, or other such remediation strategy) could help us measure the effects of these factors on discharge improvement, and lessons can be learned as data sets are generated with more frequent sampling from underground mines.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The factors that appear to complicate our prediction capacity are the ongoing changes in the underground mine including collapse of coal pillars left in the mine to support the roof, the potential for the creation of fresh faces of pyrite that can generate more acid products, ever changing flow paths with blockages and mine pools forming at different places and at different times based on seasonal flow, and variability of oxygen and pyrite within the mine over short distances. Additionally, it appears that coal barriers or seals between mines can leak or completely break thereby allowing additional flow and acidity to be introduced to an adjacent mine. Such factors as these are very difficult to measure and account for in our prediction of mine drainage improvement.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Related articles in JEQ:
This article has been cited by other articles:
|
|
 |
|
  T. S. Rotting, R. C. Thomas, C. Ayora, and J. Carrera Passive Treatment of Acid Mine Drainage with High Metal Concentrations Using Dispersed Alkaline Substrate J. Environ. Qual., August 8, 2008; 37(5): 1741 - 1751. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Grandy, G. P. Robertson, and K. D. Thelen Do Productivity and Environmental Trade-offs Justify Periodically Cultivating No-till Cropping Systems? Agron. J., October 3, 2006; 98(6): 1377 - 1383. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||
