JEQ Journal of Natural Resources and Life Sciences Education
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 ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Stewart, B.R.
Right arrow Articles by Jackson, M.L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stewart, B.R.
Right arrow Articles by Jackson, M.L.
Agricola
Right arrow Articles by Stewart, B.R.
Right arrow Articles by Jackson, M.L.
Related Collections
Right arrow Water Quality
Right arrow Water Pollution
Right arrow Industrial Waste
Journal of Environmental Quality 30:1382-1391 (2001)
© 2001 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORT
Waste Management

Evaluation of Leachates from Coal Refuse Blended with Fly Ash at Different Rates

B.R. Stewarta, W.L. Daniels*,b, L.W. Zelaznyb and M.L. Jacksonc

a Dep. of Plant and Soil Science, Mississippi State Univ., Mississippi State, MS 39762
b Dep. of Crop and Soil Environ. Sciences, Virginia Polytechnic Inst. and State Univ., Blacksburg, VA 24061-0404
c Dep. of Forestry, Virginia Polytechnic Inst. and State Univ., Blacksburg, VA 24061-0138

* Corresponding author (wdaniels{at}vt.edu)

Received for publication November 30, 1999.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
There is great interest in returning coal combustion products to mining sites for beneficial reuse as liming agents. A column study examined the effects of blending two coal fly ashes with an acid-forming coal refuse (4% pyritic S). Both fly ashes were net alkaline, but had relatively low neutralizing capacities. One ash with moderate alkalinity (CRF) was bulk blended with coal refuse at 0, 20, and 33% (w/w), while another lower alkalinity ash (WVF) was blended at 0, 5, 10, 20, and 33% (w/w). The columns were leached (unsaturated) weekly with 2.5 cm of simulated precipitation for >150 wk. Where high amounts of ash alkalinity (>20% w/w) were mixed with the coal refuse, pyrite oxidation was controlled and leachate pH was >7.0 with low metal levels throughout the study. At lower rates of alkalinity loading, trace metals were sequentially released from the WVF ash as the 5, 10, and 20% treatments acidified due to pyrite oxidation. Lechate metals increased in proportion to the total amounts applied in the ash. In this strongly acidic environment, metals such as Mn, Fe, and Cu were dissolved and leached from the ash matrix in large quantities. If ash is to be beneficially reused in the reclamation of acid-producing coal refuse, the alkalinity and potential acidity of the materials must be balanced through the appropriate addition of lime or other alkaline materials to the blend. Highly potentially acidic refuse material, such as that used here, may not be suitable for ash/refuse codisposal scenarios.

Abbreviations: AMD, acid mine drainage • CCE, calcium carbonate equivalence • EC, electrical conductance • NP, neutralization potential • PA, potential acidity • CRF, Clinch River fly ash • WVF, Westvaco fly ash • ICP-OES, inductively coupled plasma optical emission spectroscopy • USBOM, U.S. Bureau of Mines • SD, standard deviation • SEM, scanning electron microscopy


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
COAL REFUSE MATERIALS are generated during coal cleaning activities and frequently generate large areas of potentially acidic waste disposal fills (Stewart and Daniels, 1992). The use of fly ash in the reclamation of coal refuse has been the subject of several studies (Adams et al., 1972; Jastrow et al., 1981), but has never become a widespread practice in the USA. However, large amounts of coal fly ash are being bulk codisposed with coal waste materials in the Appalachian states of West Virginia and Kentucky. Conceptually, alkaline coal combustion waste products should be useful in offsetting the acid generating potential of coal wastes and acidic mine spoils. In other countries the codisposal of fly ash and coal refuse is a common practice (Skarzynska, 1995; Twardowska, 1990). Increasing landfill costs have led to greater interest in returning coal ash to mining districts for disposal. Provisions that call for coal producers to provide disposal of coal ash are being written into many coal contracts. These provisions provide an opportunity for the beneficial use of coal ash in the reclamation of coal refuse.

Leaching columns have been used by many researchers (Bradham and Carrucio, 1990; Perry, 1985; Stewart et al., 1997; Kazonich and Kim, 1997) to study acid mine drainage (AMD) generation. Leaching tests are the only controlled means to study the kinetics of pyrite weathering and alkalinity production. Both Bradham and Carrucio (1990) and Perry (1985) concur that leaching column tests give the best approximation of field weathering conditions. In a column study of surface mine spoils, Hood and Oertel (1984) estimated that each week of a continuous leaching cycle in their leaching study was equivalent to approximately 3 yr of natural weathering. This study was designed as a follow-up to earlier work reported by Stewart et al. (1997), which showed that alkaline ash blended at relatively high rates (20 and 33% w/w) appeared to prevent AMD breakthrough for multiple years. The data reported here are part of a much larger and more comprehensive column study with two fly ash sources and other more conventional (topsoil capping, blending with ground limestone, or rock phosphate) AMD treatments for comparison. Detail on those treatment effects is given by Stewart (1996) and will be presented in a future paper. In addition this study differs from the previous study in that larger diameter + length columns were packed with a larger amount of material, and the columns were designed to limit air penetration from the bottom. One ash (WVF-Westvaco) used in this study was blended at rates of 5, 10, 20, and 33% (w/w) to study the effects of varying alkaline loading to offset acidity generated from pyrite oxidation. The 20 and 33% blending rates represent amounts of fly ash that would likely be returned to the coalfields estimated based on current clean coal refuse/fly ash ratios. The 5 and 10% were included to examine the effects of placing fly ash in an environment underloaded with alkalinity, a scenario that was specifically not reported by Stewart et al. (1997) and is critically important to the field application of fly ash: refuse codisposal. A second more alkaline ash (CRF-Clinch River) was also blended at 20 and 33% (w/w).

Significant questions exist regarding the use of fly ash as an amendment in the reclamation of acidic coal refuse piles, particularly the long-term water quality effects of this practice if large amounts of fly ash are incorporated into the disposal fill. The objectives of this study were (i) to determine the effects of bulk blending coal refuse with alkaline coal fly ash on net leachate quality, and (ii) to relate these effects to the possible environmental ramifications of this practice.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The coal refuse used in this study was collected at the A.T. Massey Coal Company, Elk Run processing plant near Sylvester, WV. The refuse consisted of material that was primarily cleaned from the Peerless seam of the upper Pennsylvanian system. The refuse was air-dried, sieved to <1.9 cm (0.75 inch) and homogenized with a rotary mixer. The potential acidity (PA) of the refuse was determined by the H2O2 oxidation method of R.I. Barnhisel and J. Harrison (unpublished report, 1976, Kentucky Agric. Exp. Stn., Lexington, KY).

Two sources of fly ash were used in the study. One fly ash (WVA) was collected from the ash landfill at the Westvaco Corporation's papermill in Covington, VA. Coal from the Elk Run site is burned at the Covington Mill. Before use, the CRF ash was air-dried and homogenized using a rotary mixer. The other ash (CRF) was shipped in barrels from American Electric Power's Clinch River Plant near Carbo, VA. This ash was gathered directly from the hoppers below the electrostatic precipitators and required only homogenization with the rotary mixer. Some distilled water was sprayed on this ash to control the dust while mixing. The neutralization potential (NP) of the ash materials were determined by method no. 955.01 of the AOAC (1990). These results are expressed as calcium carbonate equivalence (CCE).

The refuse and the ashes were subjected to a total dissolution (Lim and Jackson, 1982) with subsequent elemental analysis by Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) with a Thermal Jarrell-Ash Atomscan 2400. The elemental content of the ash materials, and refuse were determined at the U.S. Bureau of Mines (USBOM), Albany, OR Research Center (Dewey, 1995). The USBOM also performed a macroscopic and microscopic fabric analysis of selected samples of post experiment column materials.

Column Design
The columns used in this experiment were constructed from smooth bore, 20.3 cm diameter ABS plastic drainage pipe, and an endcap, which was perforated with several 8 mm holes to allow for leachate drainage. The endcap was lined with 60-mesh (0.25-mm) nylon sieve cloth and Whatman no. 42 filter paper to retain the fine material. A 25.4-cm (top diam.) HDPE funnel was attached to the bottom of the column with silicone sealer. The funnel was packed with glass wool to wick the water away from the bottom of the column. To prevent gas exchange at the bottom of the column, the funnels were bonded to the column with silicone sealer. The bottom of the funnel was plugged with a 100% silicone foam rubber stopper with a Nalgene HDPE tube between the stopper and a 0.6-m length of Tygon tubing. The Tygon tubing was clamped and plugged with an HDPE stopper. This procedure allowed the leachates to be collected in the funnel portion of the column with little chemical change between sampling. The funnels held nearly 1 L of leachate, and the liquid level remained below the level of the column bottom in the funnel, so the bottom of the column was unsaturated. During the study, 5 to 10 mL of leachate was allowed to remain in the Tygon tubing to act as a gas trap. This leachate remained at the low point of the Tygon tubing loop beneath the column and prevented ambient oxygen from affecting the bottom of the column.

Column Assembly
All columns contained 36 kg of refuse to establish a constant mass of reactive material. The treatments were mixed with the refuse utilizing a rotary mixer on an oven-dry weight percent basis (Table 1). The rotary mixer was thoroughly cleaned between treatments. After all of the material required for one treatment had been mixed, that mix was weighed into column batches. Three replicate columns of each treatment were then packed. Three replications per treatment were also used by Watzlaf (1992) and Stewart et al. (1997) in experiments using smaller diameter columns. Before packing, 750 g of acid-washed glass shot was spread evenly across the nylon mesh at the bottom of each column to promote even drainage. The mixed material was precisely placed into a column using a 900-mL plastic cup suspended from two strings (Stewart, 1996). The treatments were added in 5- to 10-kg lifts that were packed to uniform density using a 3-kg baseball bat.


View this table:
[in this window]
[in a new window]
  		Table 1. Descriptions of treatments and abbreviations used in this study.

 
Leachate Collection and Analysis
Each column was inoculated with about 3 mL of AMD, to ensure an active population of pyrite oxidizing bacteria. The columns were dosed with 1 L of pH 4.6 simulated acid rain (Halvorsen and Gentry, 1990) per day until leachate broke through the bottom of the column. The columns were then put on a once-weekly watering schedule. The watering regime consisted of dosing with 2.54 cm of simulated acid rain on a weekly basis (1320 mm of rainfall per year). Initially, leachate was collected 1 d after dosing, then 2 d, then 4 d, and then after 7 d for the final 2 yr of the experiment. The effects of these changes in collection time appear to have been negligible on the basis of pH and electrical conductance (EC) data. When leachate was collected, all leachate with the exception of the 5 to 10 mL of gas trap was collected and the volume recorded. Before the next dosing, each funnel was purged of all leachate, and this volume of leachate was also recorded before the leachate was discarded. It was assumed that an airtight seal was maintained around each funnel and the leachate in the funnel was not exposed to gas exchange.

After collection, leachates were analyzed for pH and EC within 24 h (McLean, 1982; Rhoades, 1982, respectively). Samples were then preserved with trace metal grade HNO3 for subsequent analysis. Samples were analyzed for B, Cu, Fe, Mn, and S by ICP-OES. Weeks 1 through 12 were analyzed individually. Further analysis for Weeks 16 to 95 was done on 5-wk composites. Data from Weeks 103, 122, and 143 represent individual samples for those weeks. The analysis of Cu was added after the first 12 wk of the study. The first leachates were collected 24 Feb. 1992, and the columns were maintained for >150 wk total run time. The volume of leachate that passed through the column at 1, 3, 10, 21, 45, 69, 93, and 165 h after dosing was measured in May 1995.

In the first few weeks of the study, we experienced problems with algal growth in the funnels and tubing. The funnels were painted black in an effort to block out light and limit the algae growth. A component in the spray paint used caused the funnels to weaken and crack, and painting was discontinued. The funnels were then wrapped with aluminum foil to prevent light penetration. After wrapping, algal growth was not a problem. Funnel cracking was also experienced in some columns later in the experiment. When cracking occurred, leachate, and therefore data, were not collected from a column for one or two leaching cycles.

Data Analysis
Means and standard deviation (SD) for each parameter tested were calculated by treatment. These data where then plotted against time for comparison purposes. The error bars on the graphs represent 1 SD above and below the mean. Nonoverlapping error bars were interpreted as two statistically different treatments. In most instances, however, treatment effects discussed as different in this paper were at least 3 to 4 SD apart.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The refuse was tested using the H2O2 method (R.I. Barnhisel and J. Harrison, unpublished report, 1976) and found to have a potential acidity (PA) of 130 Mg of CCE demand per 1000 Mg of refuse. This test agreed well with the Leco furnace sulfur content of 4.0% S, assuming 1% pyritic-S requires approximately 31.25 Mg CCE per 1000 Mg material tested for complete neutralization (Sobek et al., 1978). This sample of refuse contains more S than much of the refuse from the coalfields of southern West Virginia and southwest Virginia and therefore represents "worst case" acid leaching conditions for this region.

As with most fly ash from bituminous coals, the NP values of these materials are relatively low and this limits their utility as liming agents. The CRF ash had a relatively high NP of 11% CCE, while the WVF material had an intermediate value of 5% CCE. These NP values were in agreement with the water pH values measured for these ash materials. The CRF had a 1:1 water pH of 11.04, while the pH of WVF was 8.02.

Aluminum, Si, Fe and C made up the bulk of the refuse and ash materials (Table 2). The amount of Ag, As, Cd, and Hg in each of the materials was at or below the detection limit. The WVF ash contained seven times more total B than the CRF ash. The WVF ash also contained higher levels of Mn than the CRF ash. The additional Mn may be attributed to wastes from a charcoal process being burned along with the coal at this plant. The WVF ash also contained higher levels of Cr, Pb, and Zn than the other materials.


View this table:
[in this window]
[in a new window]
  		Table 2. Elemental content of the refuse and fly ash materials used in this experiment; some elements are reported on a g kg-1 oxide basis (Dewey, 1995).

 
A petrographic microscope analysis of the materials performed by the USBOM (Dewey, 1995) found that WVF contained much higher levels of C than the CRF ash, which reflects the efficiency of the burning conditions at the plants. The CRF burning conditions are assumed to be hotter based on the dominance of slag beads (solid spheres) in the ash (~75% by volume based on optical estimate) and the general transparency of the beads. The WVF ash had a much higher unburned coal fragment content (70% by volume, optical estimate) and a much lower slag bead content (~20%). In addition, the beads were found to be opaque, which indicates a lower temperature burn. This may also be responsible for the higher Cr, Pb, Mn, and Zn content of the WVF ash. In addition to the slag beads, the CRF ash contained ~20% entrained unreacted lithic fragments (shale), which were composed primarily of quartz and plagioclase feldspar. The WVF ash contained coal, slag beads, and ~10% unreacted lithic fragments, of which 50% were quartz. The coal refuse was found to consist of ~50% low grade shale fragments, ~30% angular shards of coal, and ~20% primarily quartz crystals.

Examination of column materials at the conclusion of the study by optical microscopy and scanning electron microscopy (SEM) analyses confirmed that pyrite was oxidizing to form Fe-sulfates (Dewey, 1995). The probable oxidation product identified was copiapite , a ferrous–ferric sulfate that is a common oxidation product of pyrite (Nordstrom, 1982). Most of the oxidation taking place was found to be in direct association with pyrite grains. A lack of Fe-oxides proximal to the pyrite was noted. This was significant in that, if the pyrite were oxidizing to an oxide mineral (e.g., limonite or hematite) phase, it would be expected that solid phase would be found close to the pyrite. The lack of oxide coatings on the pyrite also indicated that the retardation of pyrite oxidation observed in the ash-treated columns was not due to the presence of coatings. Evangelou (1995) has presented methodology for the microencapsulation of pyrite through the formation of stable Fe-oxide coatings on pyrite. These coatings inhibit O2 diffusion to the pyrite surface.

Rate of Leachate Elution from Columns
The various treatments had significant effects on the volume of leachate eluted with time. In a typical dosing cycle, all the ash-blended treatments produced <100 mL of leachate in the first hour after leaching, while the unamended refuse produced >100 mL of leachate. The effects of increasing ash rates on leachate elution were readily observed in the WVF ash blends. The refuse columns passed leachate rapidly, while the 33% WVF and 20% WVF treatments had a delayed leachate elution. The 20% WVF treatment had a slower leachate release than the 33% WVF treatment. This did not concur with the findings of Albuquerque (1994), who found that hydraulic conductivity of refuse ash blends decreased consistently as the amount of ash in the blend increased. This trend was also observed in the data from the 20% CRF and 33% CRF treatments and was attributed to bulk density differences. The 33% treatments were not packed as tightly as the 20% treatments, and the 33% columns were 25 cm longer than the 20% treatments. This packing difference was due to the fact that at the 33% ash blending rate, the coarse refuse fragments began to float in the ash matrix. In all likelihood, this blending rate represents the highest feasible ash blending rate.

The reduction in leachate elution rate could best be described as being directly correlated with the amount of material added in the blend. When no material was blended with the refuse, leachate moved more rapidly. When 5 to 15% (w/w) material was blended with the refuse the elution of leachate was delayed, and the effects were more pronounced with time. When blending rates >20% were used, very little leachate was emitted in the first hour after dosing and an elution peak in the 10- to 21-h range was observed. Regardless, all leachates from all treatments were sampled and analyzed on the same day within a given week.

pH Effects
The effects of the various treatments on leachate pH were dramatic. The pH of the refuse treatment was initially 4.5 and decreased with each successive leaching until a pH of about 1.7 was attained after 10 wk (Fig. 1). The pH of the leachate from the refuse treatment remained at this low pH for the duration of the experiment, although it did increase slightly to pH 1.8 by the end of the experiment. The rapid nature of the pH decrease suggests that the biological oxidation of iron pyrite was active. Biological oxidation of pyrite is much more rapid than direct (O2) oxidation (Evangelou, 1995; Singer and Stumm, 1970). The initial leachate collected from the 5% WVF columns had a pH of 8, but this treatment rapidly acidified (Fig. 1). Based on the potential acidity of the refuse and NP of the WVF ash, the 5% WVF treatment was underloaded with respect to alkalinity by 128 Mg CCE per 1000 Mg of refuse. The 5% WVF treatment had some inhibitory effect on the onset of acidification, however, and delayed the pH drop by about 6 wk. The amount of acidity produced by pyrite oxidation subsequently overwhelmed the alkalinity present and the leachate pH decreased to pH 1.8 (Fig. 1). This treatment (5% WVF) subsequently maintained a pH, which was slightly higher than that for the refuse control.



View larger version (41K):
[in this window]
[in a new window]
  		Fig. 1. Chronologic mean (n = 3) column leachate pH values for each treatment. Each week 2.5 cm of simulated acid rain was applied to each column. Error bars reflect 1 sample SD above and below the mean.

 
The 10% WVF maintained a pH around the 7.8 range for the first 35 wk of the experiment (Fig. 1). This treatment was underloaded with alkalinity by 125 Mg CCE per 1000 Mg of refuse, but contained twice the alkalinity of the 5% treatment and delayed the onset of acidification twice as long as the 5% WVF treatment. The postacidification leachate equilibrium pH maintained around 2.0, which was slightly higher than the previously discussed treatments. The 20% WVF treatment maintained a pH of 7.9 through the first 100 wk of the experiment. This treatment then slowly acidified until reaching an equilibrium pH of 2.3. During the acidification process there was a large amount of variability among the replicates in this treatment. Variability during column acidification in the early phases of this experiment was also observed by Jackson (1993). The amount of variability decreased as the treatment systems approached a new equilibrium pH. The 33% WVF treatment maintained a pH of 8 throughout the experiment. This treatment was also underloaded with respect to alkalinity (by at least 100 Mg CCE per 1000 Mg of refuse), however, it has maintained a high pH for the duration studied.

The CRF-blended treatments contained higher amounts of alkalinity and exhibited similar plots for pH with time (Fig. 1). The initial leachate pH values from the CRF treatments were in excess of 9. These pH values decreased initially to around pH 8 before increasing back up to around pH 9 for the 20% CRF treatment, and to 8.4 for the 33% CRF. The difference in these pH values may have been due to the lower rate of leachate elution observed in the 20% CRF treatment (Fig. 1). Overall, these treatments maintained alkaline (pH ~ 8) leachates, and the CRF treatments clearly delayed the onset of pyrite oxidation and acidification for the duration studied, but it is unclear if this will be a permanent effect over longer periods of time.

There are several mechanisms that alone or in combination may explain the high pH maintained by the CRF ash treatments. At neutral to alkaline pH, O2 is thought to be the primary oxidizer of FeS2 (Singer and Stumm, 1970). This is due to the low solubility of Fe(OH)3 and hence low availability of Fe3+ at neutral and higher pH values. Oxidation of pyrite by O2 is known to be much slower than oxidation by Fe3+ (Singer and Stumm, 1970). By maintaining a high pH, the rapid oxidation of pyrite by Fe3+ is avoided. If pyrite oxidation does occur, there is sufficient alkalinity present to neutralize any acidity. Also, lowering hydraulic conductivity slowed the removal of reaction products (Fe, SO4, acidity) and reduced the rate of oxidation. Adding ash would decrease the pore size in refuse and may limit air infiltration and slow down gas exchange. Fly ash materials are known to have high water holding capacities (Chang et al., 1977), and the water would be held in pores, further decreasing gas exchange. Fly ash has also be shown to adsorb Fe from solution (De and Lal, 1990) and that mechanism may also serve to retard pyrite oxidation by scavenging Fe3+ from solution.

Electrical Conductivity
The electrical conductivity of a solution gives an indirect measure of the amount of dissolved ions in that solution (Rhoades, 1982). Ash materials are known to contain freely soluble salts and can produce EC values as high as 6 S m-1 (Page et al., 1979). Pyrite oxidation also produces high EC values (Evangelou, 1995; Caruccio and Geidel, 1978), due primarily to sulfate loading. The effects of both these can be seen in the data from the varying rates of WVF blended with the refuse (Fig. 2). The initial EC of the 33% WVF treatment was 0.7 S m-1, and that value gradually decreased to 0.23 S m-1 as the salts were being leached from the ash in this blend. The 20% WVF and 10% WVF treatments had slightly higher EC values than the 33% treatment initially, and then generated lower values until the onset of acidification by pyrite oxidation. Less ash was applied in these treatments and hence less salts were available for leaching, which explains the lower initial EC values. The initial increase in EC may have been due to flushing of entrained salts that had accumulated in the refuse before the columns were packed. As pyrite oxidation took place, reaction products (Fe, SO4, and acidity) were brought into solution and the EC increased. The 5% WVF and refuse treatments had higher EC values than the previously mentioned treatments, and the 5% WVF did have consistently lower EC values than the refuse treatment. The ash applications, even at the low 5% rate, limited EC via the retardation of pyrite oxidation. These treatments did not contain sufficient alkalinity to prevent pyrite oxidation, but the intensity of pyrite oxidation was decreased in these treatments when compared with the control. It is unclear whether this retardation was due to decreased oxygen penetration (Hammack and Watzlaf, 1990) or binding of Fe by the ash (De and Lal, 1990). The binding of Fe would produce effects similar to treatment with Rock-P, in which pyrite oxidation is not stopped, but the elution of reaction products is reduced (Spotts and Dollhopf, 1992). The refuse treatment produced a leachate with an EC of 2.7 S m-1. The EC decline in these treatments after acidification may be due to a decrease in the reactivity of the pyrite (Caruccio and Geidel, 1978) and flushing of reaction products. Not all pyrite oxidizes at the same rate and it seems plausible that the most reactive pyrite would be the first to be oxidized. As the most reactive material is consumed, the rate of oxidation may be limited by the reactivity and/or available surface area of the pyrite.



View larger version (40K):
[in this window]
[in a new window]
  		Fig. 2. Chronologic mean (n = 3) column leachate EC values for each treatment. Each week 2.5 cm of simulated acid rain was applied to each column. Error bars reflect 1 sample SD above and below the mean.

 
Elemental Constituents of Leachate
The leachate Fe content provides a measure of pyrite oxidation, although some Fe is most likely entrained in secondary precipitates within the columns. The refuse treatment produced the highest leachate concentration of nearly 16000 mg Fe L-1 (Fig. 3). This was approximately 7000 mg L-1 higher than next highest leachate Fe content produced by the 5% WVF. Increasing rates of WVF produced a definite effect on the peak amount of Fe in the leachate. The peak Fe content of the 5% WVF treatment was approximately half that of the unamended refuse, and the 10% WVF treatment had a peak Fe concentration, which was one-half that of the 5% WVF treatment. The 20% treatment was not fully acidified when the experiment was terminated. The leachate Fe content of the 33% WVF treatment peaked at 0.5 mg L-1. This leachate was pH 8 where Fe is quite insoluble (Stumm and Morgan, 1981). The refuse treatment eluted 662 g of Fe during the experiment, or 39% of the total Fe in the material (Fig. 4). The 5% WVF treatment eluted 520 g Fe, or 30% of the total Fe present. The 10 and 20% WVF treatments eluted 13 and 1.3% of their total Fe contents, respectively. The CRF ash leachate Fe content was slightly higher than the WVF treatments but still <1.0 mg L-1. As expected, the Fe content of the leachates was directly related to pH. Those treatments in which the pH was maintained >7 had very low Fe contents. Those WVF treatments which acidified maintained low leachate Fe contents until acidification took place, and then the Fe content increased.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3. Variation in treatment mean (n = 3) leachate Fe content with time. Each week 2.5 cm of simulated acid rain was applied to each column. Error bars reflect 1 sample SD above and below the mean.

 


View larger version (40K):
[in this window]
[in a new window]
  		Fig. 4. Mean (n = 3) total Fe (left axis) and Mn (right axis) for the WVF treatments and Control.

 
Manganese was more concentrated in the ash, particularly in the WVF ash, than in the refuse (Table 2). Other researchers have found a variety of other elements to be concentrated in fly ash (Dreesen et al., 1977) when compared to their source coal. The behavior of ash-concentrated elements under acid leaching conditions were of particular interest in this study. Manganese is usually an element of concern in AMD due to relatively high water treatment costs. The refuse treatment released Mn as it acidified and the peak Mn content was 48.5 mg L-1 (Fig. 5). After the peak elution, leachate Mn content decreased rapidly and was <5 mg L-1 by the end of the experiment. The 5% WVF treatment reached a peak leachate Mn content of 200 mg L-1. The ash (5%) in this treatment contained 0.99 g of Mn, and this treatment eluted an additional 1.2 g of Mn when compared with the control (Fig. 4). The 5% WVF treatment eluted 47% of the total Mn during the experiment, while the refuse treatment eluted only 31% of its total Mn. The 10% WVF treatment produced a peak Mn content of >460 mg L-1 15 wk later. The ash in this treatment contained 2.4 g of Mn, and an additional 2.0 g of Mn leached from this treatment (Fig. 4) compared with the refuse treatment. As with the 5% WVF treatment, 47% of the total Mn leached during the study. The 20% WVF treatment did not produce the sharp Mn peak of the 5% WVF and 10% WVF treatments, (Fig. 5) and had more within-treatment variability. The applied ash added 5.3 g of Mn, and this treatment eluted 3.6 g of Mn more than the control. This treatment had a slightly higher fraction of the total Mn leached (50%), however, and it should be noted that this treatment was actively acidifying when the experiment was terminated. Manganese elution preceded and peaked before the elution of Fe in the blended ash treatments that acidified (Fig. 5). The 33% WVF treatment eluted a very small quantity of Mn (0.004 g). These combined treatment responses are all clearly pH/solubility related effects, and the higher alkalinity CRF treatments had very low leachate Mn levels (Fig. 5).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 5. Variation in treatment mean (n = 3) leachate Mn content with time. Each week 2.5 cm of simulated acid rain was applied to each column. Error bars reflect 1 sample SD above and below the mean.

 
The Mn data illustrate an important finding, that trace metals contained in coal fly ash can become mobile in an acid leaching environment in oxidizing coal refuse. Acid mine drainage contains significant amounts of dissolved trace metals, but the metal content can actually be enriched as AMD comes in contact with fly ash at low pH. To stabilize the trace metals in the ash, the bulk acid–base long term balance in the weathering mixture must be maintained.

Copper, like Mn, was more concentrated in the ash than in the refuse (Table 2). Copper release was similar to Mn release, but did not display the sharp peaks seen in the Mn data (Fig. 6) and was lower in overall leachate concentration. Copper release reached a peak and then gradually decreased, whereas Mn displayed sharp peaks that rose and fell rapidly. The leachate from high alkalinity treatments contained low levels of Cu (Fig. 6). At pH values above 7, Cu solubility is reported to be controlled by tenorite (CuO) (Fruchter et al., 1990). The Cu levels eluted by the refuse and lower %WVF blends would clearly be of regulatory and aquatic toxicity concern.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 6. Variation in treatment mean (n = 3) leachate Cu content with time. Each week 2.5 cm of simulated acid rain was applied to each column. Error bars reflect 1 sample SD above and below the mean.

 
The leachate S was assumed to be in the form of SO4. Two types of S release curves were observed in this study: (i) in which S content was initially near 2000 mg L-1, which increased after pyrite oxidation increased, peaked, and gradually decreased; and (ii) in which the initial leachate S content was >2000 mg L-1 and the S content gradually declined to between 600 and 800 mg L-1 S. The first curve was typical of the treatments that received no alkalinity or where the alkalinity added was insufficient to control pyrite oxidation. The second type of curve was typical of the high alkalinity treatments (20 and 30% ash). The S and EC data were very similar in shape, which corroborates sulfate as the major component in the EC values. The high alkalinity treatment leachates initially had S contents ranging between 1200 and 3400 mg L-1. These high sulfate levels suggested that a sulfate salt more soluble than gypsum (CaSO4·2H2O) was controlling the S solubility in these treatments. Copiapite and other complex Fe-sulfates are likely. With time, these salts were leached away and the equilibrium S value suggested that gypsum was the solubility controlling phase. These findings concur with Garavaglia and Caramuscio (1994), who reported S solubility to be controlled by gypsum in a study of alkaline ash in lysimeters.

Leachate B content was of concern in the ash treatments because fly ash is known to contain high levels of soluble B (Bhumbla et al., 1993). Leachate B levels were directly related to the amount of B applied. Those treatments which contained 33% WVF had the highest leachate B contents (Fig. 7). The CRF ash had less B than the WVF ash and leachates from the CRF treatments exhibited correspondingly lower B levels. The apparent leachate B content in the control was confounded due to very high Fe levels before Week 40. Between the time the Week 40 and 45 samples where run, the ICP-OES was overhauled with new optics, software, and computers. All the treatments that had high Fe levels exhibited decreased B values after this point.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 7. Variation in treatment mean (n = 3) leachate B content with time. Each week 2.5 cm of simulated acid rain was applied to each column. Error bars reflect 1 sample SD above and below the mean.

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The treatments in which high amounts of alkalinity were mixed with acid-forming coal refuse (CRF treatments, 33% WVF) controlled pyrite oxidation and maintained a leachate with high pH and low metal levels throughout this 150-wk study. There are many mechanisms that potentially explain this effect. Pyrite oxidation is known to be slower at high pH because biologically mitigated oxidation occurs only at low pH. Chemical mechanisms dominate pyrite oxidation at high pH, and if direct pyrite oxidation is limited, the alkalinity in these treatments can neutralize the acidity as it is produced. Fly ash also fills the macropores of the refuse and has been shown to decrease hydraulic conductivity of refuse. It is postulated that ash blends may also decrease the movement of air into these columns, especially if the blend is kept in a fairly moist state. Metal sorption by ash has also been documented, and that mechanism could serve to retard pyrite oxidation by scavenging Fe3+, limiting its pyrite oxidation capacity. It is important to note that despite the high ash blending rates used, these columns were still underloaded with alkalinity with respect to the potential acidity of the refuse. The fact that several of the higher ash rate blended columns appeared to be entering an initial phase of acidification in the final weeks of the monitoring period may indicate that the fly ash treatment simply suppresses the onset of pyrite oxidation, but does not prevent its onset over long (>2 yr) periods of time.

The results of the varying rates of WVF application point out some of the potentially hazardous consequences of exposing ash to an intensely acidic leaching environment. Trace metals were released from the ash as the 5, 10, and 20% WVF treatments acidified due to pyrite oxidation. The amounts of metal leached increased proportionately with the total amounts applied in an ash-bound form. In this strongly acidic environment, metals such as Mn, Fe, and Cu were dissolved and leached in large quantities. This research did not address the ultimate fate of eluted metals in a refuse pile environment, but we would expect attenuation to some extent downgradient as the leachates migrate through the pile. However, if the leachates continue to migrate through highly acidic zones, we would expect many of these metals to remain in solution.

Overall, if coal fly ash is to be beneficially reused in the reclamation of acid-producing coal refuse, the alkalinity and potential acidity of the materials in the codisposal zone must be balanced. Also, attention should be paid to whether or not fly ash disposal zones are placed downgradient from acid-forming materials in a given fill. For most ash materials this will require the addition of additional alkalinity; ground agricultural limestone is a likely additive. Refuse material, such as that used here, with very high potential acidity is not suitable for ash codisposal scenarios.


   ACKNOWLEDGMENTS
 
This research was supported by Virginia Tech's Powell River Project, the A.T. Massey Coal Co., American Electric Power, the Westvaco Corp., Consol, and Virginia's Center for Innovative Technology.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





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 ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Stewart, B.R.
Right arrow Articles by Jackson, M.L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stewart, B.R.
Right arrow Articles by Jackson, M.L.
Agricola
Right arrow Articles by Stewart, B.R.
Right arrow Articles by Jackson, M.L.
Related Collections
Right arrow Water Quality
Right arrow Water Pollution
Right arrow Industrial Waste

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome