This chapter presents the application and interpretation of Acid-Base Accounting (ABA) to prediction of mine drainage quality. ABA is one mechanism for assessing postmining water quality. Its utility is amplified when used in conjunction with other premining information including baseline water quality, examination of adjacent and previous mining, and evaluation of geologic and hydrologic conditions. This chapter presumes collection, preparation, and analysis of samples following the approach suggested in Chapters 5 and 6 (Overburden Sample Collection and Preparation, and Laboratory Methods for Acid-Base Accounting). The concepts presented here are drawn largely from research and experience with ABA in Pennsylvania, supplemented, where appropriate, with information from other sources.

Acid-Base Accounting was developed at West Virginia University by Richard M. Smith and coworkers (Skousen et al., 1990). The approach grew from early attempts at classifying mine spoils for revegetation potential, based principally on acidity or alkalinity, and rock type. From these broad classifications, the need for lime and suitability for plant species could be assessed.

In 1971, Richard Smith and associates reported on the acid producing potential of rocks associated with the Freeport and Kittanning seams in Preston County, West Virginia (West Virginia University, 1971) and began to formally develop a system of balancing the acid and alkaline producing potential of rocks. Acid-base characterization work was gradually broadened to include rocks throughout the Appalachian and Interior coal basins, including Pennsylvania. The importance of acid neutralizing minerals was recognized and quantified, and the term "neutralization potential" (NP) was introduced. This work was published in a series of reports (e.g. Smith et al., 1974; 1976) and culminated in a manual of recommended field and laboratory procedures (Sobek et al.,1978). ABA in its original and modified forms has been widely adapted in both the coal and mineral mining industries in the United States, Canada, Australia and southeast Asia (Miller, 1991; British Columbia Acid Mine Drainage Task Force, 1989).

Although the potential utility as a water quality predictor was quickly recognized, the early developmental work on ABA was directed mainly toward assessing the agronomic potential of overburden and minespoil. At that time "topsoiling" (i.e. saving and reapplying topsoil) was not widely practiced and ABA was useful for identifying overburden as root zone material. The first attempts to define levels of significance for ABA data were "potentially toxic" materials having a net ABA of less than -5 ppt CaCO₃ (tons CaCO₃ equivalent/1000 tons of material), or "acid toxic" if paste pH was less than 4.0 (Smith et al., 1974, 1976; Surface Mine Drainage Task Force, 1979). These values were based on an assessment of lime requirements of native soils in the Appalachian region and plant growth needs, and not water quality conditions per se.

A first approximation was the development of a two by two matrix for qualitative interpretation of overburden analyses (Figure 11.1, Brady and Hornberger, 1990). The four fields were broadly classed into high and low sulfur and neutralization potential content. The boundaries between the four fields were not numerically defined, and the matrix served as a conceptual approach rather than explicit criteria.

(After Brady and Hornberger, 1990)

Figure 11.1 Conceptual Decision Matrix for Acid Drainage Potential Used in the Early 1980’s.

As the PaDEP and industry continued to gain experience with ABA interpretation, it became possible to focus on rock strata that were likely to be significant generators of alkalinity or acidity. These significant strata could then be evaluated in a broader context of the overall mining and reclamation plan and hydrologic and geologic conditions. Significant strata are defined by "threshold" values (Brady and Hornberger,1990) for sulfur content and neutralization potential as:

The numbers were intended for general guidance, while recognizing that rock strata with NP less than 30 or sulfur content less than 0.5 percent do influence mine drainage quality.

The PaDEP also evaluated means for representing the mass, volume, and distribution of ABA parameters within a mine site to further refine predictive capabilities. A computer spreadsheet was developed (Smith and Brady, 1990) that mass weights ABA data by strata using the Thiessen polygon method. Summary values of ABA data are calculated that adjust for the horizontal and vertical extent of each analyzed stratum. The PaDEP now also has a mainframe overburden database capable of performing and reporting the summary calculations of ABA data in various ways (Bureau of Information Services, 1993).

ABA, as originally developed, and used in Pennsylvania, consists of measuring the acid generating and acid neutralizing potentials of a rock sample. These measurements of Maximum Potential Acidity (MPA) and Neutralization Potential (NP) are subtracted to obtain a Net Neutralization Potential (NNP), or net Acid-Base balance for the rock as follows:

Net Neutralization Potential (NNP) =

NP - MPA (11.1)

The results are customarily reported in tons per thousand tons of overburden or parts per thousand. The units designation reflects the agronomic origins of ABA. An acre furrow slice of agricultural soil weighs about 1000 tons, and liming requirements are usually expressed in tons per acre (tonnes/hectare). The units of measure for ABA are therefore comparable to lime requirement designations for agricultural lands.

The components of ABA measurements are sometimes referred to by other terms, as they have been adapted for use in metal mining and other applications (Miller and Murray, 1988). The term "Acid Production Potential" (APP) is equivalent to MPA, "Acid Neutralizing Capacity" (ANC) is equivalent to NP; and "Net Acid Producing Potential" or NAPP is the same as NNP.

FeS₂ + 2CaCO₃ + 3.75O₂ + 1.5H₂O Õ

2SO₄^2- + Fe(OH)₃ + 2Ca²⁺ + 2CO₂ (11.2)

For each mole of pyrite that is oxidized, two moles of calcite are required for acid neutralization. On a mass ratio basis, for each gram of sulfur present, 3.125 grams of calcite are required for acid neutralization. When expressed in parts per thousand of overburden, for each 10 ppt of sulfur (equal to 1 percent sulfur content) present, 31.25 ppt of calcite is required for acid neutralization.

Cravotta et al. (1990) noted that the stoichiometry in Equation 11.2 is based on the exsolving of carbon dioxide gas out of the spoil system. They suggested that in a closed spoil system, carbon dioxide is not exsolved, and additional acidity from carbonic acid is generated. Cravotta et al. (1990) proposed that up to four moles of calcite might be needed for acid neutralization as follows:

FeS₂ + 4CaCO₃ + 3.75O₂ + 3.5H₂O Õ

2SO₄^2- + Fe(OH)₃ + 4Ca²⁺ + 4HCO₃^- (11.3)

The stoichiometry of Equation 11.3 shows that twice as much calcite would be required for acid neutralization. On a mass basis, for each 10 ppt of sulfur present, 62.5 tons of calcite is needed for acid neutralization in one thousand tons of overburden.

The choice of which stoichiometry most closely describes a minespoil system directly affects the ABA calculation, alkaline addition rates, and prediction of expected postmining water quality. Brady and Cravotta (1992) in analyzing ABA and water quality data from 74 mine sites, showed that correct prediction of postmining net alkalinity was improved from 52 to 57 using a stoichiometric equivalence of 62.5. They also found that the "errors" in prediction become more balanced using a 62.5 factor, with equal proportions of sites erroneously predicted to yield acid water and sites predicted to yield alkaline water. However, a later study initiated by the PaDEP (Brady et al., 1994) showed that the 31.25 equivalence factor was most accurate, correctly predicting postmining net alkalinity on 31 of 38 mines (82%), while the 62.5 factor correctly predicted 22 of 38 (58%). The diverse results from these two studies underscore the complex geochemical processes at work and unique character of each minesite. Data reduction methods may also influence site rankings.

The PaDEP’s mainframe ABA database processes overburden data using both the 31.25 and 62.5 factors. The experience base of ABA interpretation has been built largely on using the 31.25 factor, and this convention is followed in most of the subject literature.

Neutralization potential is presumed to measure carbonate minerals, exchangeable bases, and weatherable silicate minerals (Sobek et al., 1978), and provide an index of available acid neutralizers in the rock. The procedure does not discriminate among forms of neutralizers and represents a theoretical maximum value for NP.

FeCO₃ + 2H⁺ Õ Fe²⁺ + CO₂ + H₂O (11.4)

Fe²⁺ + 0.25O₂ + H₂ O + H⁺ Õ

Fe³⁺ + 1.5H₂O (11.5)

Fe³⁺ + 3H₂ O Õ Fe(OH)₃ + 3H⁺ (11.6)

Summary Reaction:

FeCO₃ + 0.25O₂ + 1.5H₂ O Õ Fe(OH)₃ + CO₂ (11.7)

The initial dissolution of siderite and ferrous iron oxidation consumes acid and may provide temporary neutralization. However, iron hydrolysis ultimately generates acidity equal to that initially consumed, resulting in no net neutralization.

Interpretation of NP data and the possible carbonate minerals present is facilitated by examination of "fizz" or reaction with room temperature hydrochloric acid. Samples with NP less than 20 ppt CaCO₃ generally will not fizz or effervesce (Sobek et al., 1978). Where present in sufficient concentrations, calcite will fizz readily, dolomite fizzes slowly and siderite will not react at ambient temperature. Thus the PaDEP’s "rule of thumb" for NP greater than 30 ppt CaCO₃ with fizz serves as a qualitative check on the carbonate mineralogy and the rock's ability to neutralize acid.

Besides carbonates and bases, the NP procedure may extract some silicate minerals which weather only slowly under field conditions. The alteration of silicate minerals can consume large amounts of acidity over periods of geologic weathering (Appelo and Postma, 1993) but are a minor source of short term acid neutralization. Inclusion of acid extractable silicates in NP can overestimate the readily available neutralizing capability.

Lappako (1994), working with metal mine wastes, found that NP was overestimated in samples containing calcium feldspar and other minerals.

Maximum potential acidity is based on a measure of sulfur content of the rock. The presumption is that this accurately represents the amount of acid generating sulfur minerals. Sulfur in overburden occurs in sulfide, sulfate, or organically bound forms (see Chapter 10). Metal sulfides, mainly pyrite, are the principal source of acid generation and the dominant sulfur form in Appalachian overburden (Smith, et al., 1976). For overburden analysis performed in Pennsylvania, the PaDEP has found MPA from total sulfur to be a reliable index of acid generating potential. Acid-base accounting stoichiometry assumes the sulfur is all present as pyrite, and complete oxidation and acid generation occurs. Thus the designation of maximum potential acidity.

Sulfate sulfur exists in many minerals such as gypsum (calcium sulfate), jarosite (potassium-iron sulfate), and alunite (potassium-aluminum sulfate). Sulfate minerals are usually present in significant quantities only in weathered spoil or refuse; and otherwise absent from fresh overburden in Pennsylvania. Alkaline earth sulfate salts like gypsum are nonacid formers. Metal sulfate salts, however, are intermediate products of pyrite oxidation, and represent "stored acidity". These minerals can undergo dissolution and hydrolysis with acid generation. Sulfate sulfur cannot be ruled out as a potential acid source unless the mineralogy is known, and the common lab procedures for sulfur fractionation do not identify the specific minerals present.

Sulfate minerals such as gypsum are common in the arid and semi-arid regions of the western United States and total sulfur may not be a reliable index of MPA. Adjustments to ABA analysis, such as sulfur fractionation, for western mine lands are discussed in Williams and Schuman (1987).

Organically bound sulfur is generally considered to be to nonacid forming and is found in coals, carbon rich shales, partings, "bone coal", etc.

Sulfur fractionation may be useful where the material analyzed is refuse or weathered spoil and significant partitioning of sulfur among the three fractions is suspected.

Computation methods for NNP imply that acid generation from MPA and acid neutralization from NP take place concurrently and at equal rates. In fact, acid generation can proceed more rapidly (see Chapter 1, Geochemistry of Coal Mine Drainage), and is catalyzed by Thiobacillus ferrooxidans bacteria. The solubilities of acid products are such that total acidity may range from hundreds to thousands of mg/l. Carbonate mineral dissolution is a function of pH and partial pressure of carbon dioxide (Plummer et al., 1978) and does not occur as rapidly. Alkalinity in mine waters is seldom greater than 400 mg/l due to the limited solubility of calcite and other carbonate minerals.

The traditional computation method for NNP utilizes MPA calculated from total sulfur times a 31.25 equivalence factor. Numerous modifications to the NNP calculation have been developed based on specific needs. Joseph et al. (1994) give a summary of various ways of calculating NNP for both hydrologic and agronomic interpretations. Some examples and applications are as follows:

NNP = NP - MPA, (11.8)

where MPA is percent total sulfur x 31.25.

The traditional method assumes all sulfur is in a potentially acid generating form, and is usually appropriate for fresh unweathered overburden in Pennsylvania.

NNP = NP - PA, (11.9)

where PA is percent pyritic sulfur x 31.25.

This method assumes sulfide sulfur is the only acid generating source; sulfate and organic sulfur are assumed to be nonacid generating. This is applicable where alkaline earth sulfate salts are present, but not applicable for metal sulfate salts.

NNP = NP - MPA, (11.10)

where MPA is percent total sulfur x 62.5

This assumes a closed system with no carbon dioxide gas exsolved where both sulfuric and carbonic acid must be neutralized.

NNP = NP - (PA + EA), (11.11)

where PA is percent pyritic sulfur x 31.25, and EA is exchangeable acidity.

This applies to soils and agronomic interpretations.

NNP = NP - (PA + SMP), (11.12)

where PA is percent pyritic sulfur x 31.25, and SMP is the lime recommendation from buffer pH.

This applies to soils and agronomic interpretations.

Calculations based on the traditional method shown in Equation 11.8 are the most common method of computation in Pennsylvania and other Appalachian states.

Paste pH was originally included by Smith and coworkers as one of the ABA parameters. It is seldom applied today in Pennsylvania, but continues in use in some other states. Paste pH shows the current acidity status of the sample but may provide little indication about the future behavior of the rock.

Postmining drainage concentration of metals and dissolved constituents such as sulfate are not directly predictable from ABA analysis. The ABA procedure quantifies mineral groups and not individual elements or chemical species. The behavior of individual metals or other dissolved constituents is best evaluated from consideration of pH, redox status, and solubility considerations.

Acid-base accounting is one of a group of interpretive tools for predicting mining impacts to hydrologic systems. The accumulated experience, developed from over 15 years of using ABA in Pennsylvania, shows that the strongest interpretations come from collective evaluation of several factors. These include ABA data, historical performance, premining water quality, stratigraphy and lithology, ground water flow systems, weathering effects, and the proposed mining and reclamation plan. A prediction of mine drainage quality results from an integrated evaluation of all of these factors. In this chapter, however, the discussion is limited to ABA data.

Almost from the inception of acid-base accounting, there have been attempts to define numerical criteria or levels of significance for classifying ABA results and expected rock behavior. These numeric criteria have taken the form of (1) boundaries on NNP values; (2) ratios of NP to MPA; and (3) boundaries on values for NP or MPA. Some of these criteria, and their geologic and geographic applications, are presented in Table 11.1.

Values in the table refer to characteristics of individual rock samples. Variation exists in the reported values, which are drawn from diverse geologic settings and climates. Some general conclusions are summarized as follows:

The lack of universal criteria is not surprising since mine drainage quality is a product of the interaction of many geologic, hydrologic, climatic, and mining factors. diPretoro and Rauch (1988) stated "Because of the complexity of coal mining and reclamation operations, professional judgment (taking into account lithology and history of drainage quality in a given area, as well as acid-base account parameters ) will always be required to arrive at a final determination." Miller et al., (1991) noted that "... the application of a single absolute cut-off criterion can be misleading and result in overdesign of the waste management requirements." For metal mines, it has been suggested that ABA criteria are site specific and mineral dependent. (Miller and Murray, 1988; Morin and Hutt, 1994).

Application of ABA has been likewise proposed either as (1) the principal overburden analysis tool (Sobek et al., 1978) or (2) as an initial screening tool. Where ABA results are unclear, simulated weathering tests, mineralogic characterization, or other analyses have been recommended (Carruccio and Geidel, 1980; Miller et al., 1994; British Columbia Acid Mine Drainage Task Force, 1989).

Dual interpretive criteria have evolved for applications to water quality predictions, and soils and revegetation concerns. Efforts in Pennsylvania and other eastern coal states have focused heavily on water quality predictions with ABA as the primary overburden analysis tool.

ABA and mine drainage quality relations have been evaluated in Pennsylvania and northern Appalachia in three studies, including projects initiated by the PaDEP, West Virginia University, and the U.S. Bureau of Mines. These studies have shown that carbonate content of the overburden or NP is a very important factor controlling mine drainage quality. In each study, net alkalinity (alkalinity minus acidity) was used as the primary index of postmining drainage quality. The parameters acidity, alkalinity, and net alkalinity are measures of the complete acidity or alkalinity generating capacity of a water. They are also the aqueous analogues of the ABA rock parameters of MPA, NP, and NNP.

The PaDEP conducted a study that included about 40 surface mines from Pennsylvania's bituminous coal field (Brady et al., 1994, and Perry and Brady, 1995). Each mine had two or more ABA drill holes and multiple postmining water quality samples from seeps, springs, or monitoring wells. Raw ABA data were processed into a summary value for the entire mine using mass weighting procedures described by Smith

Table 11.1 Summary of Suggested Criteria for Interpreting Acid-Base Accounting ⁽¹⁾

and Brady (1990). Summary ABA data were compared to median water quality values. Summary NP and NNP values were computed with and without the "threshold" criteria of 0.5 percent sulfur and NP of 30 ppt CaCO3 with fizz. Eleven different coal beds are represented from the Allegheny, Conemaugh, Monongahela, and Dunkard groups as shown in Table 11.2. Sites included single and multiple seam operations with mining methods ranging from small block cut operations to area mining by dragline.

Principal findings of the study are:

Table 11.2 Coal beds represented in Pennsylvania acid-base accounting study.

Mines with NP greater than 21 ppt CaCO₃ (mass weighted basis) all produced net alkaline water. Eight of eleven mines with NP less than 10 ppt CaCO₃ had net acid water. Five of these mines had net alkalinity of less than -200 mg/L. Mines with mass weighted NP between 10 and 21 ppt CaCO₃ included six acid and ten alkaline sites with most waters between -125 and +150 mg/L net alkalinity. A scatterplot of these data is shown in Figure 11.2.

Figure 11.2 Pennsylvania ABA-Mine Drainage Study Neutralization Potential vs. Net Alkalinity

Figure 11.3 Pennsylvania ABA-Mine Drainage Study Net Neutralization Potential vs. Net Alkalinity

Net alkalinity was also compared to NNP and NP computed with threshold criteria applied. Results for NP and NNP with and without thresholds are summarized in Table 11.3.

Figure 11.4a Plot of Cumulative Frequency of Alkaline Waters vs. Neutralization Potential

Figure 11.4b Plot of Cumulative Frequency of Alkaline Waters vs. Net Neutralization Potential

Figure 11.5 is a scatterplot of MPA versus acidity in postmining waters. There is no apparent relationship between the two, and MPA by itself is not a good predictor of mine drainage quality. Obviously, pyrite must be present for acid generation to occur, but the abundance or lack of carbonate minerals controls the overall evolution of acid-base properties of mine drainage. Smith et al. (1976), examining ABA data in the southern Appalachians, concluded that some rocks are so low in sulfur content that they are incapable of generating acid conditions regardless of their carbonate content.

Figure 11.6 Pennsylvania ABA-Mine Drainage Study Sulfate Production (Normalized) vs. Neutralization Potential

Alkaline conditions greatly limit the activity of dissolved ferric iron. Removal of dissolved ferric iron by alkaline conditions is important since it interrupts the self propagating acid cycle. Dissolved ferric iron is capable of rapidly oxidizing pyrite as follows:

FeS₂ + 14Fe³⁺ + 8H₂O Õ

15Fe²⁺ + 2SO^2-₄ + 16H⁺ (11.13)

Finally, carbonates are acid reactive, with their dissolution rate a function of H⁺ activity (pH) and the partial pressure of CO₂ (Plummer et al., 1978). As acidity increases, the rate of carbonate dissolution increases. Conversely, under alkaline conditions, carbonate dissolution slows until equilibrium is reached.

The inhibitory effects of carbonate minerals on acid generation have been observed in laboratory weathering studies (Williams et al., 1982; Carrrucio and Geidel, 1980) which showed that rocks containing several percent pyrite produced less acidity and sulfate when carbonate was present at a few percent concentration. Lorenz and Tarpley (1963) noted that in coal samples containing calcite, "calcite increased the pH of the reaction from 3.5 needed for optimum growth of Ferrobacillus ferroxidans to a higher value that inhibited the catalytic effect of the bacterium".

A study by diPretoro and Rouch (1988) of ABA and mine drainage quality in northern West Virginia included many of the same coals and stratigraphic sequences mined in Pennsylvania such as the Waynesburg and upper Freeport intervals. Their study also used mass weighting of ABA data and shows results similar to the Pennsylvania study. The findings of diPretoro and Rauch are summarized as follows:

MPA showed no apparent relationship to net alkalinity. Also, diPretoro and Rauch noted lithologic effects, in that mines with a large proportion of sandstone overburden could produce acid drainage even at low sulfur contents.

Erickson and Hedin (1988) examined ABA and water quality relations on 32 mines. About half the sites were in Pennsylvania; the remainder were in other northern Appalachian states and the Illinois basin. Their study also used volume weighted ABA data. Findings are summarized as follows:

The sample group did not have any sites with NNP between 20 and 80 ppt CaCO₃. The authors also cautioned against using ABA data alone, and suggested analyzing other information such as past water quality data.

Data reported elsewhere (Engineer's International, 1986) from the same mines in Erickson and Hedin's study showed that pH and alkalinity were generally related to NP and NNP, and specific conductance and NP were also correlated.

NP and NNP results from the three Appalachian studies are compared in Table 11.3. Variation in results among the studies may stem from differences in how ABA data were processed and summarized. The Pennsylvania study used slightly different geometric approximations. Differences may also reflect regional or stratigraphic variation in rock properties.

Other coal overburden and metal tailings studies have examined the utility and predictive capabilities of ABA against weathering tests. Sturey et al. (1982), in a comparison of ABA and column weathering studies, concluded that both methods were adequate to show acid and nonacid trends and that ABA provided more information on individual strata. Bradham and Carrucio (1991) analyzed 10 metal mine tailings samples and compared them to field weathering test pads. ABA predictions for acid or alkaline quality agreed with field performance for eight samples, while laboratory weathering tests agreed with field performance of all 10 samples. Most samples were acid producing.

Ferguson and Morin (1991) found that Canadian metal tailings with NNP < 0 ppt CaCO₃ were generally acid producing and concluded that the NP to MPA ratio was useful as a qualitative predictor. They noted that waste rock drainage quality was more difficult to predict due to heterogeneous properties of the spoil. Ferguson and Robertson (1994) reported that no rock with an NP to MPA ratio greater than 1 had been conclusively identified as producing acid in the field. However, for screening ABA data, they propose a ratio of greater than 2 as "nonacid generating", and ratios of 1 to 2 as "possibly." Patterson and Ferguson (1994) found that an NP to MPA ratio of 1 separated acid from nonacid producing metal mines in Canada and Sweden.

Plumlee et al. (1993), sampling metal mine drainages in Colorado, concluded that drainage quality was controlled by ore deposit geology, climate, and mining method. Pyrite content and acid buffering capacity were considered to be the most important geological controls on pH and metal content of the mine waters.

This section presents an abbreviated example of review and interpretation of an ABA data set from Pennsylvania, and the resultant postmining water quality.

ABA interpretation begins with review of the overburden analysis report, including drill logs and lithologic descriptions. Within each ABA drill hole, a zone of geologic weathering occurs where carbonates (NP) and sulfides (MPA) have been removed. Highly weathered rocks are not capable of generating significant alkalinity, and the lack of sulfur precludes the formation of acid drainage. The weathered zone can be considered chemically inert and contributes little to postmining drainage quality. Weathered materials are characterized by shades of brown, yellow, or red (Munsell color chromas greater than 2), and the rocks are often weakly cemented or partly decomposed. Smith et al. (1974, 1976) suggested that the weathered zone in Appalachia is about 20 ft (6 m) deep. A study in Saltlick Township Fayette County, Pennsylvania found sulfide sulfur weathered to depths of 16 to 20 ft (5 to 6 m), (Figure 8.59, Chapter 8) while carbonate leaching in Wharton twp., Fayette County was also about 20 ft (6 m) (Brady et al., 1988).

An example drill log and ABA data from southwestern Pennsylvania (Figure 11.7) illustrates the effects of weathering on sulfur and carbonate content. Weathering in this locale has proceeded to a depth of about 18 ft (5.5 m). Weathering is indicated by the soil zone, brown clay (probably weathered shale) and light brown shale. Below 18 ft (5.5 m), the rocks are described in shades of light and dark gray, rather than brown. Total sulfur values in the upper 18 ft (5.5 m) are very low (less than 0.2 percent). NP values are also quite low (less than 5 ppt CaCO₃). The upper 18 ft (5.5 m) of overburden lacks the ability to generate either significant acidity or alkalinity. At depths below 18 ft (5.5 m), both sulfur content and neutralization potential values increase. These rocks will determine the post mining water quality.

Lithologic logs of drill holes analyzed for ABA can be compared and correlated to other drill hole logs whether or not they have geochemical analyses. ABA is usually run on only a fraction of the exploratory borings drilled on a minesite. Lithologic comparisons are useful to determine the type and degree of variation in stratigraphy, depth of weathering, or structural considerations that may signal changes in overburden geochemistry. Lithologic comparisons are also useful to correlate significant acid or alkaline strata among drill holes. If these significant strata can be correlated across a mine site and are identifiable in the field,

analysis of overburden quality and mine planning can be more rigorous. If the materials cannot be clearly correlated or the materials cannot be distinguished in the field, overburden interpretation and material handling are more speculative.

Volume weighting has been used to improve the representation of the vertical and horizontal extent of strata in a mine site. This is necessitated by the hilly topography and more or less flat lying strata of Pennsylvania's coal fields which make the upper strata less extensive than the deeper rocks. The department uses the algorithms described by Smith and Brady (1990) which incorporate area, volume and density of rock, and fraction of unit spoiled to calculate mass weighted ABA data for individual samples, entire drill holes, or multiple drill holes within a mine. These data can be further refined by calculations with and without "threshold" values for sulfur and NP, stoichiometric equivalence factors of 31.25 or 62.5, inclusion of alkaline addition rates, and these can be reported in various formats.

An "area of influence" is defined for each ABA drill hole by delineating the limits of mining and dividing the area into polygons. Figure 11.8 shows a mine with five ABA drill holes, and the polygons (area of influence) drawn for each drill hole. The areas are planimetered or digitized and the weights of NP and MPA calculated. More elaborate calculations can be made with volumetric and contouring software. However, the Thiessen polygon process described by Smith and Brady (1990) is simple enough to perform on a computer spreadsheet and can be easily modified to site specific situations.

Figure 11.9 shows an example printout of the raw and calculated ABA parameters for drill hole 17. The printout provides information by sample on depth, thickness, rock type, percent sulfur, NP, unit weight of the rock, area represented, fraction of unit spoiled (portion of overburden returned to the backfill), mass of NP, mass of MPA, mass of Net NP, and mass of overburden represented. In this example, area values increase with depth, reflecting the greater areal extent of the deeper rocks, and the overburden is modeled as a truncated cone. NP and MPA values for these deeper strata are given greater weighting in the summary calculations. Smith and Brady (1990) discuss appropriate means of modeling various overburden geometries in more detail.

Figure 11.10 provides acreage weighted ABA summary numbers for the entire drill hole calculated with and without "threshold" criteria using stoichiometric equivalence factors of 31.25 and 62.5. For drill hole 17, the overburden has an overall NP of about 35 ppt CaCO₃, overall NNP of 17.8 ppt CaCO₃ and NP to MPA ratio of 2 when calculated without thresholds and equivalence factor of 31.25 (upper left box in Figure 11.10). These results show overburden that would be expected to produce alkaline drainage.

ABA data from multiple drill holes can be combined to produce summary data for the entire mined area. These summary outputs reduce a large amount of raw data to a manageable output sufficient to analyze and interpret a mine site. Drill hole 17 was combined with two other ABA holes of similar geochemistry. Summary ABA results for the combined three drill holes are shown in Table 11.5.

Overburden Analysis Data System

Acid Base Accounting

Drill Hole Data

Id: Permit: or Project Code:

Operator: Mine Name:

County: Municipality:

Quad Map: Stream Code:

Drill Hole Id: 17 Latitude: Longitude:

UTH Zone: Northing: Easting: Date Drilled 04-SEP-90

Driller:

Drilling Method: 1 AIR ROTARY

Interpreter:

Laboratory:

Added Alkaline: Threshold Values: Sulfur: 0.500 NP: 30.00 Fizz:

Intervals @ 31.25

LEGEND: (<=Less Than) (>=Greater Than) (+ = Plus) (-= Minus) (*=Degree of Error)

Coal Seam Names:

Figure 11.9 Example of Calculated Acid Base Accounting Data, Drillhole 17

Table 11.5 Example weighted summary overburden data for three drill holes.

These results are also indicative of rocks expected to produce net alkaline drainage.

Postmining drainage quality on the example mine is net alkaline (Table 11.6) as predicted from examination of ABA data. Baseline water quality from a spring is presented for comparison.

Data for both the spring and mine drainage are median values. The mine drainage has much higher alkalinity than the spring, as would be expected from the accelerated weathering of calcareous rocks in the minespoil. Both the mine drainage and the spring contain very low levels of metals. Sulfate concentrations in the mine drainage indicate that some sulfide oxidation has occurred within the minespoil. However, the production of alkalinity is sufficient to neutralize and/or inhibit in-situ acid generation.

Table 11.6 Postmining Water Quality Resulting From Mining of Overburden in Drill hole 17

Acid-Base Accounting has been applied to the prediction of overburden and water quality properties on mined land for about 25 years. An extensive institutional knowledge base and "rules of thumb" have developed on the interpretation of ABA data, and have been supplemented with a few formal studies. Carbonate or NP content exerts major control over the postmining water quality. NP contents of as little as 20 to 30 ppt CaCO₃ equivalent are effective in producing alkaline drianage. ABA is a valuable assessment and prediction tool. Its application is most effective as part of a group of interpretive techniques for analyzing the interaction of geologic and hydrologic conditions, and mining and reclamation practices.