JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lamminen, M.
Right arrow Articles by Traina, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lamminen, M.
Right arrow Articles by Traina, S. J.
Agricola
Right arrow Articles by Lamminen, M.
Right arrow Articles by Traina, S. J.
Related Collections
Right arrow Water Quality
Right arrow Heavy Metals
Right arrow Water Pollution
Right arrow Industrial Waste
Journal of Environmental Quality 30:1371-1381 (2001)
© 2001 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORT
Waste Management

Effect of Flue Gas Desulfurization (FGD) By-Product on Water Quality at an Underground Coal Mine

Mikko Lamminena, James Woodb, Harold Walker*,a, Yu-Ping Chinb, Yongtian Hea,b,c,d and Samuel J. Trainad

a Ohio State Univ., Columbus, OH 43210
b Ohio State Univ., Columbus, OH 43210
c Environmental Science Graduate Program, The Ohio State Univ., Columbus, OH 43210
d Ohio State Univ., Columbus, OH 43210

* Corresponding author (walker.455{at}osu.edu)

Received for publication June 26, 2000.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 SITE DESCRIPTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
In this paper, a field study was carried out to examine the effect of flue gas desulfurization (FGD) by-product on water quality at an underground coal mine in central-eastern Ohio. Flue gas desulfurization by-product was injected into the down-dip portions of the Roberts–Dawson mine in an attempt to seal major seeps exiting the mine and to coat exposed pyritic surfaces. Immediately following grout injection, significant increases in acidity, iron, aluminum, sulfur, and calcium were observed at most surface and ground water locations near where grouting was carried out. Following this initial flush of elements, concentrations of most constituents have decreased to near pre-grouting levels. Data from the site and geochemical modeling suggest that an increase in water level or rerouting of drainage flow resulted in the dissolution of iron and aluminum sulfate salts and ferrihydrite. Dissolution of the FGD grout material resulted in increases in calcium and sulfate concentrations in the drainage waters. Water within the mine voids was saturated with respect to calcium sulfate and gypsum immediately following grout injection. Based on an analysis of core samples obtained from the site, acid mine drainage (AMD) was in contact with at least some portions of the grout and this resulted in grout weathering. Subsequent transport of calcium and sulfate to the underclay, perhaps by fracture flow, has resulted in the deposition of gypsum and calcium sulfate solids.

Abbreviations: AMD, acid mine drainage • FGD, flue gas desulfurization • SI, saturation index • XRD, X-ray diffraction


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
SITE DESCRIPTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
THE Clean Air Act Amendments of 1990 required coal-burning electric power plants to decrease emissions of sulfur oxides to the atmosphere. To meet this objective, many power plants that use high-sulfur coal have installed wet lime (CaO) scrubbing systems. This scrubbing process produces a calcium sulfite slurry that is dewatered and mixed with fly ash and residual lime. The resulting product is usually referred to as stabilized flue gas desulfurization material, or stabilized FGD. Approximately 22.7 x 106 metric tons of FGD were produced in the United States in 1998, with 90% going to landfills (Stewart and Kalyoncu, 1999).

In order to reduce the volume of stabilized FGD entering landfills, a number of beneficial uses for dewatered calcium sulfite sludge and FGD have been developed. At present, the majority of FGD not entering landfills is oxidized to form synthetic FGD gypsum and used in the manufacture of wallboard (Stewart and Kalyoncu, 1999). Flue gas desulfurization by-product is also an effective substitute for clay in the construction of low permeability liners, animal feed lots, and hay storage pads (Butalia et al., 1999). There is increasing interest in the utilization and/or disposal of FGD at abandoned mine lands (AMLs). Typically, coal mine areas are located near electric power plants that burn high-sulfur coal and produce FGD material. Effective reuse of FGD at these sites reduces land requirements for FGD disposal and may aid in the reclamation of AMLs.

Research at surface mine sites has demonstrated that FGD can be used effectively in the reclamation of areas containing mine spoil (Dick et al., 1994; Stehouwer et al., 1995a,b; Hao et al., 1998). For example, Stehouwer et al. (1995b) examined the use of FGD material for the reclamation of acidic mine spoil in greenhouse experiments. Dry FGD was utilized to raise the pH of acidic mine spoil and enhance revegetation of ‘Kentucky 31’ tall fescue (Festuca arundinacea Schreb.). Increased plant yield was observed upon application of FGD, most likely as a result of a decrease in soluble aluminum and its associated toxicity. Decreases in As, Cd, Cr, and Se in plant tissue were also observed with FGD amendment. Boron toxicity was not observed. In addition to mine spoil reclamation, the large-scale placement of FGD in abandoned deep mines is also being considered. Successful placement of FGD in deep mines would reduce landfill requirements, and potentially aid in subsidence control and acid mine drainage (AMD) reduction. Despite the interest in the placement of FGD in underground mines, little information is available regarding the effect of FGD on water quality in AMD environments.

A recent laboratory study examined the weathering of FGD material in contact with AMD (Laperche and Traina, 1999a,b). Significant decreases in iron, zinc, manganese, magnesium, and aluminum and increases in pH, calcium, sulfur, arsenic, selenium, boron, sodium, and potassium were observed. Mineralogical analysis showed that exposure of FGD to AMD resulted in a loss of minerals such as ettringite and hannebachite and the formation of gypsum. As a result of weathering, the physical structure of the FGD grout deteriorated upon exposure to AMD. Thus, while FGD may increase the pH of AMD, it may not be an appropriate material for providing structural strength. It should be noted, however, that the water to solids ratio in these laboratory experiments may not have been representative of conditions observed in the field. In a true field situation, only a small fraction of the total surface area of the FGD material may be in contact with AMD, and therefore, the extent of weathering may be significantly lower.

In this paper, a field study was carried out to examine the effect of FGD on water quality at an underground coal mine in central-eastern Ohio. Flue gas desulfurization by-product was injected into the down-dip portions of the Roberts–Dawson mine in an attempt to seal major seeps exiting the mine and to coat exposed pyritic surfaces. Before and after grouting, extensive surface water and ground water monitoring was carried out to determine the effects of grouting on water quality. These data, along with information about the structure and properties of FGD core samples collected from the mine, are used to assess the controlling processes influencing water quality at the site.


   SITE DESCRIPTION
TOP
 ABSTRACT
 INTRODUCTION
 SITE DESCRIPTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
The study was carried out at the Roberts–Dawson mine located on the borders of Coshocton and Muskingum counties, in central-eastern Ohio (Fig. 1, Panel A). The mine spans an area of 0.059 km2 (14.6 acres) and had approximately 6 x 104 m3 of coal removed during its operation until it was closed in the 1950s. The geology of the site is shown in Panel B of Fig. 1. A detailed discussion of the geology at the site can be found elsewhere (Bair and Hammer, 1999). The Middle Kittanning (#6) coal layer is 1 to 2 m thick. This layer is just below 0 to 60 m of interbedded sandstones and shales of the Freeport Sandstone. Beneath the middle Kittanning coal is a 1- to 1.5-m-thick "underclay" that overlies 9 to 12 m of shale and siltstone. Below the Vanport Limestone lays the fine-grained Clarion Sandstone. The Freeport Sandstone and middle Kittanning #6 coal form perched water tables that overlie the regional water table within the Clarion Sandstone, approximately 18 m below the coal layer.



View larger version (42K):
[in this window]
[in a new window]
  		Fig. 1. (A) Site location of the Roberts–Dawson mine. (B) Each ground water sampling site contains a cluster of three wells; one in the upper Freeport Sandstone, one in the middle coal layer, and one in the lower Clarion Sandstone. (C) Surface water and ground water sampling locations relative to known mine voids at the Roberts–Dawson mine site. Shaded regions represent areas in the mine voids where high-strength stabilized flue gas desulfurization (FGD) was placed. Low-strength stabilized FGD was injected into the unshaded regions of the mine voids. The direction of flow in the receiving stream is indicated by a Q.

 
Between October 1997 and January 1998, FGD grout was injected into the down-dip portions of the Roberts–Dawson mine. The FGD grout used at the Roberts–Dawson site was a 1.25:1 mixture of fly ash and dewatered scrubber sludge with an additional 5% lime (CaO). The FGD grout consisted primarily of calcium, silicon, iron, sulfur, and aluminum (Laperche and Traina, 1999a,b). Other minor elements were also present including antimony, arsenic, barium, beryllium, boron, carbon, chromium, cadmium, copper, lead, manganese, nickel, potassium, sodium, selenium, strontium, and zinc. A total of 317 boreholes were drilled to inject the FGD grout into the mine voids and 18182 m3 of grout material was injected (Damian and Mafi, 1999).

Panel C of Fig. 1 shows the known mine voids at the Roberts–Dawson site. There was also a significant area of unmapped mine voids on the southern portion of the site. The shaded areas in Panel C of Fig. 1 indicate locations were high-strength FGD grout was injected into known mine voids in order to seal the main seeps. A lower-strength grout was injected into the unshaded regions of the known mine voids to coat pyritic surfaces. Flue gas desulfurization grout was also injected into limited areas of the unmapped portion of the mine. The design strength of the grout used for coating the surface of the mine voids was approximately 520 kPa (75 psi) after 91 d, while the grout used to seal the main seeps had a strength of about 1000 kPa (145 psi) after 91 d (Damian and Mafi, 1999).

Two major seeps discharging AMD were identified at the Roberts–Dawson mine site. These sites are numbered 3 and 5 in Panel C of Fig. 1. Site 5 drained largely from the mapped portion of the mine voids while Site 3 drained the unmapped portion of the mine. However, it should be noted that these two seeps were not hydraulically isolated. The seeps discharged into an adjacent receiving stream, which then flowed into a collection pond. Acid mine drainage exited the collection pond and discharged through a culvert to Wills Creek Reservoir.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 SITE DESCRIPTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
Water Quality Sampling
Extensive surface water and ground water quality monitoring was conducted at the Roberts–Dawson site both before and after grouting operations. Surface water samples were generally collected monthly from 12 sites. In this paper, water quality data for Site 5 will be discussed. This sampling location was selected because it consistently had the highest flow rates and chemical fluxes of any seep at the site. More than 30 ground water wells were installed at the project site to characterize changes in ground water quality and water level elevation. Ground water monitoring wells were placed in the upper Freeport Sandstone aquifer and the coal layer, as well as in the lower Clarion Sandstone (see Panel B of Fig. 1). Ground water sampling wells were located upgradient of the mine on the northern end of the site, as well as within the mine void region. Ground water wells were made with either 2.5-cm (1-in), 5.1-cm (2-in), or 10.2-cm (4-in) Schedule 40 PVC pipe. The minimum 5.1-cm annular well space was sealed with a bentonite grout via a tremie pipe method. The upper terminus of the monitoring wells conformed to ASTM Standard D 5787-95 (American Society for Testing and Materials, 2000).

In this paper, data for four wells (9705, 9719, 9728, and 9906) will be discussed. These wells were chosen because they consistently produced samples for analysis and were representative of distinct geologic and hydrologic locations at the site. Well 9719 (depth = 48.0 m) was located in the coal layer, near the exit of the main seep of the mapped portion of the Roberts–Dawson mine area. This site was representative of ground water downgradient of most of the mine voids, immediately prior to exiting seeps. Well 9705 (depth = 21.6 m) was also located in the coal layer, in the mapped portion of the mine voids, and was upgradient of Well 9719. This site, therefore, represented an area of the mine that experienced less cumulative effect from mine drainage. Well 9728 (depth = 27.5 m) was located on the western side of the site in an area containing acidic mine spoil. Thus, the geology of this immediate area is distinctly different from the known mine voids. Well 9906 (depth = 17.4 m) was installed following the retrieval of an FGD core sample on the northwest portion of the site, and was the only location where grout mineralogy data was also available. Although only a subset of sites is presented, it should be noted that the trends observed in these wells and surface water locations are consistent with those seen at other, similar locations at the site.

Flow rates at the surface water sites were measured using weirs or by the bucket and stopwatch technique. The bucket and stopwatch technique was used if flow rates were low enough to obtain accurate readings using a 15-L bucket. If flow rates were high, the flow was recorded based on the water level in a weir and the calibrated flow reading for that water level. All reported flow data represent the average of three independent measurements. Surface water samples were collected using a disposable 60-mL Luer-Lok syringe (Becton Dickinson, Franklin Lakes, NJ) and placed in 60-mL polypropylene bottles. A total of four samples were collected at each surface water site. Two of the samples were placed directly into a polypropylene bottle and the remaining two samples were filtered on site with a 0.45-µm disposable sterile cellulose acetate membrane filter (Corning, Wilmington, NC). Sample bottles, syringes, and filters were thoroughly rinsed with sample water prior to sample collection. A new syringe and filter was used for every sample collected. Sample bottles were filled completely to minimize the amount of oxygen present. Filtered and unfiltered samples for metals analysis were acidified to a 10% (v/v) acid concentration using ultrapure nitric acid. Following collection, samples were immediately placed in an iced cooler for transport back to the laboratory.

The water levels in ground water wells were measured with a Heron (Hamilton, OH) water level probe. Prior to sample collection, at least three volumes of water were removed from the well. Wells were purged using either dedicated submersible Redi-Flow pumps (Ben Medows Co., Canton, GA) or by using a Reel (Redmond, WA) E-Z portable well pump. Well pumps were controlled using a Grundfos (Clovis, CA) BMI/MP1-115V pump controller and powered with a 3.7 kW (5 hp) generator (Briggs and Stratton, Milwaukee, WI). Some wells were purged manually using disposable 1-L, high-density polyethylene bailers (Timco Manufacturing, Prairie du Sac, WI).

Water Quality Analyses
All samples were analyzed for pH, conductivity, sulfate, arsenic, chloride, alkalinity, metals, and other inorganic constituents. We measured pH using a Model 525A pH meter (Thermo Orion, Beverly, MA). The pH electrode and meter were calibrated prior to each analysis using either a two-point or three-point calibration procedure. On some occasions, pH was measured in the field and compared with laboratory measurements carried out at a later time. The pH measurements carried out in the laboratory were generally within 0.1 pH units of pH measurements conducted in the field. Conductivity was measured in the laboratory using a digital conductivity meter (Fisher Scientific, Suwanee, GA). The conductivity meter was calibrated prior to analysis using known specific conductance standards.

Alkalinity, chloride, and sulfate were determined using a Lachat (Milwaukee, WI) Quickchem AE autoanalyzer. For alkalinity, chloride, and sulfate analyses, laboratory blanks were periodically checked using ultrapure deionized water and compared with a laboratory blank control chart. One duplicate sample was analyzed for every 20 samples. Check standards were analyzed for every 30 samples. For alkalinity determinations, hydrochloric acid was used and standardized using potassium hydrogen phthalate (KHP). For chloride determinations, a mercuric thiocyanate ferric nitrate solution was used as a color reagent and chloride standards were prepared using analytical-grade sodium chloride. For sulfate determinations, standard solutions were prepared using reagent-grade potassium sulfate dried at 110°C for at least 2 h. If sample concentrations were above the range of standards, samples were diluted and reanalyzed.

Arsenic was determined using a PerkinElmer (Norwalk, CT) Model 4100XL graphite furnace atomic absorption spectrometer. Analyses for Al, Ba, Be, B, Cd, Cr, Ca, Co, Cu, Fe, Pb, Li, Mg, Mn, Mo, Ni, P, K, Si, Na, Sr, S, and Zn were carried out using an inductively coupled plasma atomic emission spectrometer at the Ohio Agricultural Research and Development Center (OARDC) in Wooster, OH. Laboratory blank samples were run using ultrapure deionized water. One laboratory blank was run for every 20 samples. One duplicate was run for every 10 samples by splitting a sample. Instrument check standards were prepared separately from instrument calibration standards and run once for every 10 samples. The error in analyses was less than 5% based on duplicate samples. The percent difference in the anion–cation balance was generally less than 10%.

Acidity in units of mg/L as CaCO3 was calculated based on the following formula:

[1]
where the concentrations of iron, aluminum, manganese, and protons are in mol/L. The constants within the parentheses are the number of equivalents per mole for each species. The equivalent weight of calcium carbonate used in Eq. [1] above is 50000 mg/eq.

Grout Core Analysis
Grout core samples were collected from the site approximately 1 and 2 yr following the grouting operation. Upon collection in the field, core samples were immediately transported to Ohio State University for mineralogical analysis by X-ray diffraction (XRD). The XRD analysis was carried out using an X-ray diffractometer (Phillips Analytical, Natick, MA) using Cu K{alpha} radiation at 35kV and 20 mA. Measurements were made using a step scanning technique with a fixed time of 4 s/0.05°2{Theta}, from 8 to 55 or 60°2{Theta}. Prior to analysis, grout core samples were air-dried and ground with a synthetic sapphire mortar and pestle to <=250 µm. Crystalline phase assignments were made based on comparative analyses of reference samples, searches of the International Center for Diffraction Data (ICDD), and data in the published literature. Following collection of grout cores, water quality wells were installed at these locations. Water quality from one grout hole (Well 9906) in which a significant quantity of FGD was retrieved will be discussed in this paper.

MinteqA2 Calculations
Water chemistry before and after grouting was modeled using MinteqA2 Version 4.0 (USEPA, 1999) to determine the saturation indices (SI) of potential solid phases. Saturation indices were determined as SI = log IAP/K, where IAP is the ion activity product and K is the equilibrium constant. For all calculations, the original database supplied with MinteqA2 Version 4.0 was used without modification. This version includes corrections to the thermodynamic database of Version 3.0 (Serkiz et al., 1996). All MinteqA2 simulations presented were performed at a temperature of 25°C at a fixed pH value corresponding to the field samples. The elements considered in the calculations included Al, Ca, Fe, Mg, Na, Mn, S, and K. Ionic strength was taken into account using the Davies equation, which is sufficient for ionic strength values less than about 0.5 M. The ionic strength was typically less than 0.06 M, except for the sampling dates in December 1997 and January 1998, in which the ionic strength values were 0.08 and 0.12 M, respectively. Charge balances were calculated for each simulation and were generally within 10%. These calculations assume equilibrium conditions within the mine voids and coal layer. Solid phases with SI values less than 0 are considered undersaturated. In this case, dissolution is thermodynamically favored over precipitation. For solid phases with SI values greater than 0, the solution is supersaturated and precipitation is favored thermodynamically.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 SITE DESCRIPTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
SUMMARY AND CONCLUSIONS
 REFERENCES
 
Prior to grout injection, water quality at the Roberts–Dawson site had the characteristic signature of AMD, with low pH and high levels of metals and other trace elements. All water samples from ground water wells located in the coal layer and at the main seeps (Sites 3 and 5) were severely affected by AMD on all sampling trips prior to grout injection. The concentrations of various constituents at Site 5 are shown for a sampling date prior to FGD grout injection (13 Mar. 1997) in Table 1. As can be seen, the pH was very low (pH = 3.3) and there were high levels of dissolved iron (24.9 mg/L), aluminum (6.0 mg/L), and sulfate (129 mg/L). Water in the receiving stream above the main seeps and in the upper Freeport Sandstone was relatively unaffected by AMD (data not shown).


View this table:
[in this window]
[in a new window]
  		Table 1. Concentrations of various constituents at Surface Water Site 5 for select sampling dates before and after flue gas desulfurization (FGD) grout injection. Concentrations are in mg/L unless otherwise noted.

 
Surface and Ground Water Quality after Grout Injection
Grouting of the Roberts–Dawson mine began in October 1997 and continued until the beginning of January 1998. Water quality data for Site 5 on 10 Jan. 1998, just at the end of the grouting operations, is shown in Table 1. Although the pH increased slightly to 3.69, significant increases in acidity, sulfate, arsenic, aluminum, calcium, iron, sulfur, and other elements were observed. For example, calcium increased from 70 mg/L on 13 Mar. 1997 to 302 mg/L on 10 Jan. 1998, a roughly fourfold increase. Sulfate increased from 129 to 1182 mg/L over the same period. More than a 10-fold increase in dissolved iron was observed, with the iron concentration increasing from 24.9 to 271.8 mg/L. These values represented some of the highest concentrations recorded at Site 5 over the entire period of the sampling program. It should be noted that sampling on 10 Jan. 1998 occurred shortly after a large rain event. Over the three days preceding sampling (7–9 Jan. 1998), 8.23 cm (3.24 in) of rain and melted snow were recorded at the Coshocton Climatological Observing Station.

Nine and eighteen months following the end of grout injection, measured levels of acidity, iron, sulfur, calcium, and other elements decreased from values recorded in January 1998 (Table 1). However, the levels of these constituents at this particular site remained higher than values observed prior to grout injection. Moreover, pH decreased to levels below pre-grout conditions. It should be pointed out that significant variations in concentrations of elements in the coal layer were observed following grout injection, perhaps due to changes in flow levels and other seasonal factors at the site. Water quality in the Freeport Sandstone, surface water sites located upstream of the main seeps, and in Wills Creek Reservoir were not affected by the grouting program (data not shown). A ground water well located in the Clarion Sandstone, directly below the known mine voids on the southeastern side of the site, indicated that the regional aquifer at this location was not affected by AMD following grouting operations.

To better understand the trends in water quality, flow rate and water quality data for Site 5 are shown in Fig. 2 for select parameters (acidity, iron, calcium, sulfur, and aluminum). The shaded region on each graph represents the period in which grouting with FGD took place. The line in each graph represents the average pre-grout level or concentration. For this particular site, drainage emerged from the ground approximately halfway between the main exit of the original seep and the receiving stream following grouting. As a result, the contribution of flow from Site 5 to the receiving stream remained relatively unchanged after grouting, with the exception of some seasonal variability (see Fig. 2). Immediately following grouting, acidity, iron, sulfur, calcium, and aluminum from this new seep increased significantly and remained elevated until March or April of 1998. Since this time, the levels of acidity, iron, sulfur, calcium, and aluminum have steadily decreased and are approaching pre-grout concentrations. Because flow to the receiving stream has not been significantly reduced, the increases in constituent levels following grouting cannot be attributed to a concentration effect. More likely, the grouting procedure altered flow paths in the mine voids, which rerouted water to previously inaccessible areas. As a consequence, the dissolution of mineral oxides and salts and/or the formation of new AMD by iron sulfide oxidation contributed iron, sulfur, and other elements to the present seep.



View larger version (53K):
[in this window]
[in a new window]
  		Fig. 2. Water quality at Site 5 before and after grouting with stabilized flue gas desulfurization (FGD) material. The line in each figure represents the average pre-grout concentration for the element or parameter. The shaded region represents the period of grouting.

 
Data in Fig. 3 and 4 show water quality parameters in ground water wells located in the Kittanning #6 coal layer. Water quality in Well 9705 before and after grout injection is shown in Fig. 3. This well was located in the coal layer in the mapped portion of the mine voids. Following grouting operations, the water level in this well increased by at least 0.3 m and up to 1 m, depending on the particular sampling date. The levels of acidity, iron, sulfur, calcium, and aluminum increased significantly immediately following grouting operations. After this initial increase, the quantities of all parameters shown in Fig. 3 (except water level) have generally decreased with time. Currently, the levels of acidity and iron are approximately equal to pre-grout levels. The concentration of sulfur and calcium, however, has remained slightly elevated compared with pre-grout levels. Conversely, the amount of aluminum has significantly decreased compared with pre-grout conditions. The pH in this well has decreased from 7.26 in June 1997 to 4.25 in June 1999.



View larger version (50K):
[in this window]
[in a new window]
  		Fig. 3. Water quality in Well 9705 before and after grouting with stabilized flue gas desulfurization (FGD) material. The line in each figure represents the average pre-grout concentration for the element or parameter. The shaded region represents the period of grouting.

 


View larger version (51K):
[in this window]
[in a new window]
  		Fig. 4. Water quality in Well 9719 before and after grouting with stabilized flue gas desulfurization (FGD) material. The line in each figure represents the average pre-grout concentration for the element or parameter. The shaded region represents the period of grouting.

 
Well 9719 was also located in the coal layer but near the exit of the main seep (Site 5) of the mapped portion of the Roberts–Dawson mine area. Water quality in this well before and after grout injection is shown in Fig. 4. After grouting operations, water levels in Well 9719 have generally been above the average pre-grout level. On only three sampling trips were levels recorded below the pre-grout average. The water levels in this well have been seasonal with the highest water levels typically observed during the winter and spring. Immediately following grout injection large increases in acidity, iron, sulfur, calcium, aluminum, as well as other elements were observed. However, levels of all constituents shown in Fig. 4 have steadily decreased since the maximum concentrations were observed in December 1997, similar to the behavior observed for Well 9705. Concentrations of aluminum have reached pre-grout levels. The levels of acidity, iron, sulfur, and calcium remain slightly elevated as compared with average pre-grout levels. The fact that peak concentrations of most elements were observed in December 1997 for this well, instead of later in January or February like some other sites, may be due to the fact that grouting took place first in the down-dip portions of the mine area near this well. The pH in this well remained relatively constant between June 1997 (pH = 3.52) and June 1999 (pH = 3.83).

Well 9728 was located within the mine spoil on the southwestern side of the site. Figure 5 provides water level and water quality data for Ground Water Well 9728. Water quality in this well was very acidic, had low pH, and high levels of iron, manganese, sulfur and other inorganic constituents prior to grout injection. Water levels following grouting have generally been higher than pre-grout levels. During the wettest months water levels were close to 1.5 m (5 ft) higher than pre-grout average levels. In addition, water quality in Well 9728 has improved significantly since grouting operations. The highest levels of acidity, iron, sulfur, calcium, and aluminum were observed just prior to or during grout injection. Since this time, levels of all constituents shown in Fig. 5 have steadily decreased and are currently well below pre-grout levels. The pH in the this well, however, has decreased from 6.28 in August 1997 to 4.10 in June 1999.



View larger version (49K):
[in this window]
[in a new window]
  		Fig. 5. Water quality in Well 9728 before and after grouting with stabilized flue gas desulfurization (FGD) material. The line in each figure represents the average pre-grout concentration for the element or parameter. The shaded region represents the period of grouting.

 
Chemical Controls on Water Quality
The surface and ground water data in the coal layer and seeps provide a consistent picture of the effect of grouting on water quality at the Roberts–Dawson mine. For virtually all surface water and ground water sampling locations affected by the grouting operations, a sharp increase in the concentrations of acidity, iron, sulfur, calcium, and aluminum were observed immediately following grouting. Following this initial peak, levels of most elements have approached or reached pre-grout concentrations. A notable exception to this behavior is the water quality trend observed in the spoil layer on the western side of the site. Here, significant improvements in water quality occurred following grouting with FGD.

A number of chemical processes may influence water quality at the Roberts–Dawson site, such as (i) metal salt dissolution, (ii) metal sulfide oxidation, (iii) dissolution of FGD grout material, and (iv) ion exchange reactions. Nordstrom and coworkers have shown that at pH values below about 4.5, the dissolved concentrations of iron and aluminum in AMD may be controlled by evaporated iron and aluminum sulfate salts such as alunite [KAl3(SO4)2(OH)6], jurbanite [Al(SO4)(OH)·5H2O], siderotil [FeSO4·5H2O], coquimbite [Fe2(SO4)3·9H2O], and basaluminite [Al4(SO4)(OH)10·5H2O] (Nordstrom, 1982). For example, it has been demonstrated by Nordstrom that the infiltration of rain water and inundation of mine voids can result in the dissolution of these aluminum and iron sulfate salts. A similar phenomenon may have occurred at the Roberts–Dawson mine as a result of FGD grout injection. Injection of FGD grout probably resulted in the rerouting of water within the mine voids and coal layer. This rerouting of flow may have exposed metal oxides and evaporated metal sulfate salts to water, and subsequently led to metal dissolution and the observed high levels of dissolved iron, aluminum, and sulfate following grout injection. Following this initial "flush," dissolved iron, aluminum, and sulfate levels would be expected to decrease, again as observed at the Roberts–Dawson site.

Metal sulfide oxidation may also play a role in controlling water quality at the Roberts–Dawson site following grout injection. The oxidation of pyritic material depends on the available oxygen content and the concentration of ferric iron (Nordstrom and Ball, 1986; Stumm and Morgan, 1981; Moses et al., 1987). The FGD grout injection program surely decreased the amount of exposed pyrite in the immediate vicinity where grouting took place. Further, it has been shown that the CaSO3 in FGD can scavenge oxygen and ferric iron, thus reducing the potential for AMD pyrite oxidation and AMD formation (Hao et al., 1998). However, rerouting of flow into previously inaccessible areas could provide new exposed surfaces for AMD formation by pyrite oxidation. An increase in the amount of ferric iron as a result of the dissolution of ferrihydrite, jarosite, or ferric salts could also facilitate pyrite oxidation. This phenomenon may explain the observed decrease in pH at Site 5, after a brief initial increase, following the grouting operation. The initial increase in pH may have been the result of FGD lime dissolution. Once this initial "titration" occurred, production of new AMD by pyrite oxidation resulted in a subsequent decrease in pH.

The effect of flow rerouting on pyrite oxidation may explain the decrease in iron, aluminum, calcium, and sulfur levels observed in the spoil layer following grout injection. Near this location, grout injection may have rerouted a significant amount of AMD-containing ground water away from the spoil layer. As a result, a greater portion of the flow through this layer would be due to relatively pristine waters from the northwest side of the site. The lower levels of ferric iron in these waters would significantly reduce the potential for pyrite oxidation and result in improvements in water quality. The fact that both sulfur and calcium have decreased in the spoil layer suggests that this location was not hydraulically connected to mine waters in contact with FGD. Therefore, the improvements in water quality in Well 9728 are probably not a result of neutralization reactions resulting from dissolution of lime from FGD material.

Previous laboratory studies have shown that AMD may accelerate the dissolution of FGD grout and result in the release of significant quantities of metals and other inorganic constituents (Laperche and Traina, 1999a). The FGD used at the Roberts–Dawson site consisted primarily of calcium, silicon, and sulfur, with lesser amounts of aluminum, iron, and other trace elements. As shown earlier, large increases in calcium and sulfur were observed immediately after grout injection. Significant increases in silicon were also observed immediately after grouting at Surface Water Site 5 and ground water monitoring wells located where grouting took place (data not shown). These data suggest grout dissolution may contribute to the elevated levels of calcium and sulfur observed at the site. A similar observation was made at the Winding Ridge site in Maryland in which injection of FGD grout (atmospheric fluidized bed combustion ash) into an abandoned mine also resulted in increased levels of calcium and sulfate (Rafalko and Petzrick, 1999). At this site, the elevated levels of calcium, and at least a portion of the sulfate loading, was attributed to grout dissolution. As a result, both grout dissolution and the dissolution of metal salts and/or pyrite oxidation probably contribute to the observed concentrations of sulfur in ground water and surface water at the Roberts–Dawson site.

The FGD grout injected into the Roberts–Dawson mine contained a significant amount of lime. Despite the acid neutralization capacity (ANC) of the FGD material, no significant decrease in acidity, and only a small initial increase in pH, was observed at monitoring locations immediately after grouting occurred. This lack of neutralization immediately after grouting may be a result of the increased acidity produced by the dissolution of metal sulfate salts and subsequent precipitation of iron and aluminum hydroxide solids. Precipitation of iron and aluminum hydroxide solids onto the surface of the FGD grout may influence the release of hydroxide ions into solution. In fact, laboratory batch experiments have demonstrated that iron hydroxide coatings form on the surface of grout material upon exposure to AMD (Wood and Chin, 1998; Walker et al., 1999). Batch experiments using AMD from the Roberts–Dawson site and a mineral acid control (HNO3, pH = 3.3) showed relatively slow increases in pH in the presence of grout material. After 10 d, both control and AMD experiments resulted in high equilibrium pH values of ~10.2 and 9.5, respectively. What differed was the response of the grout after the initial solutions were removed and replaced by fresh HNO3 or AMD. In the former scenario, the pH rebounded to pH 9.0 after 12 d. Conversely, in the experiments with Roberts–Dawson AMD, the grout ANC was severely affected, and was unable to raise the AMD pH beyond 4.0 after 12 d. The presence of iron oxide precipitates in the AMD experiments suggests that their formation altered the ANC properties of the FGD grout. These results are consistent with field observations, where a slight increase in pH followed by a pH decrease was observed. Indeed, if new AMD production occurred in the mine voids, the initial exposure of the grout to preexisting AMD would have "short-circuited" its ability to neutralize fresh AMD formed when waters were rerouted to new pyritic surfaces in the mine shaft.

The rerouting of water within the mine voids due to grouting could also lead to the leaching of ions from surrounding soil by an ion exchange mechanism or through soil dissolution processes. A recent study demonstrated that both ion exchange and soil dissolution can contribute to the release of calcium, magnesium, sodium, and potassium in the presence of acidic coal pile leachate (Zelmanowitz et al., 1995). At the Roberts–Dawson site, increased levels of dissolved sodium were observed in Ground Water Well 9719 immediately following grout injection (see Table 1). These high levels of sodium may have been caused by the preferential exchange of calcium for sodium from the surface of highly weathered clays. The dissolution of calcium from the FGD grout would enhance this exchange mechanism as a result of the high dissolved calcium levels. The increased acidity of the mine drainage immediately following grout injection may also facilitate the dissolution of calcium-bearing minerals such as calcium carbonate and calcium-containing clays in the subsurface.

MinteqA2 Analysis
To develop a better understanding of the roles of different solids phases in controlling the solution chemistry at the Roberts–Dawson site, thermodynamic calculations were carried out using MinteqA2. These calculations were performed assuming thermodynamic equilibrium, a temperature of 25°C, and that all elements were in their most oxidized state. While equilibrium conditions may not prevail, these calculations provide information regarding the relative importance of different solid phases in controlling solution composition. Saturation indices for calcium sulfate, gypsum, AlOHSO4, ferrihydrite, and jarosite are shown in Fig. 6 for Well 9719. For the calcium-containing solids (i.e., gypsum and CaSO4), an increase in SI was observed during and immediately following injection of FGD grout. In fact, SI values for both calcium sulfate and gypsum were close to zero in December 1997, near the end of grouting. This suggests that dissolved calcium levels at this time were controlled by these solid phases. MinteqA2 simulations performed at lower temperatures did not significantly affect this conclusion. It is interesting to note that slight decreases in the saturation indices of calcium sulfate and gypsum were observed in January and April of 1999. These data were collected during or immediately following storm events in which levels of calcium and sulfur were diluted due to enhanced infiltration of rainwater.



View larger version (34K):
[in this window]
[in a new window]
  		Fig. 6. Saturation indices for CaSO4, gypsum [CaSO4·2H2O], ferrihydrite [Fe(OH)3], jarosite [KFe3(OH)6(SO4)2], and Al(OH)SO4 as a function of sampling date for Well 9719. All calculations were performed using MinteqA2 Version 4.0.

 
A number of iron and aluminum solids were also found to have SI values near or greater than zero. However, no clear trend in SI values for these solids was observed over the course of the sampling program. These data support the idea that grouting resulted in the release of iron and aluminum from compounds present at the mine site prior to the start of the grout injection program. For jarosite K, SI values were orders of magnitude above saturation. The log K value used in MinteqA2 Version 4.0 for jarosite K is -14.8. Filipek et al. (1987) suggest, however, that the solubility of jarosite K may be much greater. In fact, they note that the log K value determined by Chapman et al. (1983) of -9.21 may still underestimate jarosite K solubility. Nevertheless, these thermodynamic calculations indicate that iron hydroxides and evaporated metal sulfates play a significant role in controlling the dissolved concentrations of iron and aluminum at the Roberts–Dawson site. This conclusions is in line with previous studies that have shown that AlOHSO4, jarosite K, and ferrihydrite all play major roles in controlling iron and aluminum solubility in acid mine drainage environments (Nordstrom, 1982; Filipek et al., 1987). It should be noted, however, that ferrous sulfate solids may also be important (Filipek et al., 1987), though MinteqA2 calculations for the Roberts–Dawson mine did not reveal supersaturated solid phases containing ferrous iron, even when all of the dissolved Fe was modeled as Fe(II).

Grout Core Analysis
Grout core samples were collected from the site approximately 1 and 2 yr following the initial grouting program. Mineralogical analysis was carried out on these samples as well as on samples collected from the site prior to grout injection. Mineralogical analysis of core samples from the site approximately 1 yr following grouting demonstrated the presence of gypsum in the underclay material. In fact, visible deposits of gypsum crystals were observed. Mineralogical analysis of the underclay prior to grout injection, however, showed no evidence for the presence of any calcium sulfate solids. These data are in line with the MinteqA2 analysis, which indicated that gypsum and calcium sulfate were above saturation levels immediately following grout injection and could potentially precipitate. However, it is not clear how significant levels of calcium and sulfate could migrate to the underclay and form gypsum given the low permeability of this material. One hypothesis is that fractures within the underclay formed as a result of previous mining operations, and this facilitated the transport of calcium- and sulfate-enriched AMD and the formation of gypsum in this geologic layer following grout injection.

Samples of FGD grout collected from the site 2 yr after grouting showed XRD-detectable ettringite [Ca6Al2 (SO4)3(OH)12·26(H2O)] and hannebachite [2CaSO3· (H2O)] mineral phases commonly associated with the hardening of FGD materials (Fig. 7). However, the pH of the core sample containing ettringite had a pH level of around 9.5, well below the level typically required for ettringite formation, and at the low end of the ettringite thermodynamic stability field (Myneni et al., 1998). This suggests that the interior of the grout was not in contact with AMD during ettringite formation, perhaps as a result of the low permeability of the grout material. Following the initial formation of ettringite, weathering of the FGD grout by AMD may have occurred at the Roberts–Dawson site. This weathering process would result in a decrease in the pH of the grout, the continued release of calcium and sulfur, and the eventual deterioration of the grout material. It should also be noted that the consistency of the grout core samples was similar to a paste, and therefore had significantly lower strength than samples of grout tested in the laboratory. This indicates that weathering of the grout may indeed by occurring, or that the strength of the grout never developed in the environment of the mine voids. The XRD analysis also confirmed the presence of ferrihydrite. Although iron and aluminum sulfate solids were not observed in this core sample, these solids may be important at other locations within the mine voids.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 7. X-ray diffraction pattern of flue gas desulfurization (FGD)-grout core recovered from the Roberts–Dawson mine 2 yr after injection. The formula of ettringite is Ca6Al2(SO4)3(OH)12·26(H2O) and the formula of hannebachite is 2CaSO3·(H2O).

 
Following the collection of grout cores, water quality wells were installed at these locations and water samples collected in order to better understand the water chemistry in the immediate vicinity of the FGD. Water quality data for Grout Hole 9906 is shown in Table 2. Grout Hole 9906 was located on the northwest side of the site (as shown in Fig. 1). Water collected at this site had high pH, high conductivity, and relatively low levels of Al, Fe, Mn, and S. Levels of Ca were similar to values obtained at other Roberts–Dawson sampling sites prior to grouting operations, while the amount of sodium was higher. Beryllium, Co, Cu, Ni, P, and Zn were all below the limit of detection. Examining these data using MinteqA2 indicated that this water was saturated with respect to ferrihydrite but not calcium sulfate, gypsum, AlOHSO4, or jarosite solids. These data suggest that ground water within the FGD grout cores may have been neutralized by the FGD material. The fact that high levels of sodium were measured in Well 9906 may be indicative of dissolved species from the FGD exchanging with monovalent ions in the soil matrix. It should be noted, however, that Grout Hole 9906 was not screened solely in the FGD layer, and therefore ground water from the upper aquifer not affected by AMD may also have contributed water to this well. Also, water quality data prior to grouting at this location were not available, and therefore the quality of mine drainage prior to grouting was not known.


View this table:
[in this window]
[in a new window]
  		Table 2. Concentrations of various constituents at Grout Hole 9906. Concentrations are in mg/L unless otherwise noted.

 

   SUMMARY AND CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 SITE DESCRIPTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
REFERENCES
 
Important processes controlling water quality at the Roberts–Dawson mine area are shown schematically in Fig. 8. Upon injection of FGD grout, an increase in water level or rerouting of mine drainage probably resulted in the dissolution of iron and aluminum sulfate salts, and also possibly the dissolution of ferrihydrite. Dissolution of the grout material contributed calcium and sulfate to the drainage waters, and perhaps localized neutralization of AMD. After an initial flush of constituents, levels of most elements rapidly approached pre-grout levels. In the spoil layer, flow of AMD-containing ground water may have been diverted by the grouting operations. As a result, these ground water wells experience a greater contribution from surrounding alkaline base flow and subsequently have improved water quality.



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 8. Conceptualization of processes controlling water chemistry at the Roberts–Dawson mine site following injection of flue gas desulfurization (FGD) grout.

 
Based on the core analysis, AMD is in contact with portions of the grout and this may be resulting in grout weathering. Transport of calcium and sulfate to the underclay, perhaps by fracture flow, has resulted in the deposition of gypsum and calcium sulfate solids. Thermodynamic modeling indicated that immediately following grout injection ground water at the site was saturated with ferrihydrite, iron, and aluminum sulfate solids, as well as gypsum and calcium sulfate, indicating that these solid phases play an important role in controlling water chemistry.


   ACKNOWLEDGMENTS
 
The authors would like to thank Doug Beak in the School of Natural Resources at Ohio State University for carrying out the arsenic, alkalinity, sulfate, and chloride analyses. We also thank Kevin Jewell at OARDC in Wooster, OH for performing the ICP–AES analysis. This project was funded in part by the Ohio Coal Development Office, Ohio Department of Development, under OCDO Grant no. D-95-17 and the Ohio Environmental Protection Agency. Additional support was provided by American Electric Power, Ohio Department of Natural Resources, U.S. Department of Energy, Dravo Lime Company, Office of Surface Mines, Corp of Engineers, U.S. Environmental Protection Agency, and The Ohio State University. Partial salary support was provided by OSU/OARDC.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 SITE DESCRIPTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
C.-M. Cheng, H. W. Walker, and J. M. Bigham
Mechanisms Controlling the Leaching Kinetics of Fixated Flue Gas Desulfurization (FGD) Material under Neutral and Acidic Conditions
J. Environ. Qual., May 7, 2007; 36(3): 874 - 886.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
M. K. Shukla, R. Lal, and M. H. Ebinger
Physical and Chemical Properties of a Minespoil Eight Years after Reclamation in Northeastern Ohio
Soil Sci. Soc. Am. J., June 28, 2005; 69(4): 1288 - 1297.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lamminen, M.
Right arrow Articles by Traina, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lamminen, M.
Right arrow Articles by Traina, S. J.
Agricola
Right arrow Articles by Lamminen, M.
Right arrow Articles by Traina, S. J.
Related Collections
Right arrow Water Quality
Right arrow Heavy Metals
Right arrow Water Pollution
Right arrow Industrial Waste

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
T