Lecture No. 21. Pit Lakes II

Although acid is often considered the primary water-quality problem in pit lakes, sulfate and dissolved metals and metalloids also can pose serious problems, even in neutral lakes. The secondary USEPA drinking water standard for sulfate allows a maximum concentration of 250 mg/L, which is often exceeded by an order of magnitude in pit lake water. Essential nutrient metals such as iron, copper, manganese, zinc, and cobalt, as well as nonessential elements such as lead, cadmium, arsenic, and mercury in pit lake water can exceed concentrations safe for aquatic life or domestic, industrial, or agricultural use. Arsenic can be an especially troublesome contaminant in many cases, because it is quite soluble at neutral or alkaline pH, and because it is toxic to humans and wildlife at concentrations well below 1 mg/L. An example of a neutral lake contaminated by arsenic was the Summer Camp Pit at Getchell, Nevada. The water in this pit was nearly neutral but typically contained arsenic at levels between 1 and 5 mg/L (cf. the USEPA maximum contaminant level for arsenic in drinking water, 50 m g/L). High arsenic is common in epithermal and mesothermal gold deposits of the Carlin Trend and related ore bodies in the Basin and Range; one of the defining properties of a "Carlin-type" gold deposit is the presence of arsenic. Since carbonate rock is also common in Carlin-type deposits, a pit lake that forms in such a setting is likely to be near pH 7 but contain high concentrations of arsenic.

Yet some pit lakes exist that have good water quality. In some cases the reason is a lack of available sulfide minerals and/or large amounts of carbonate in the host rock. For example, a number of coal pits in Montana, Wyoming, and the Dakotas were investigated by Anderson and Hawkes, who compared their water quality to that of bentonite mining pits and livestock watering ponds in the same area. They found no significant differences among the waters from the different sources. Both alkaline ground water and the very low level of sulfur in and around the coal seams were probably responsible for the generally good water quality in the coal pits. Coal from the northern Great Plains averages just 0.6% sulfur, much of which is in the form of refractory organosulfur compounds. There is another set of coal-mine pit lakes with fair water quality in Illinois, some of which meet drinking water standards. One such lake has been used as a drinking water source by the town of Astoria, IL, since 1976. The wall rock and overburden around most of these lakes contain limestone. The Cortez Pit, a former gold mine in eastern Nevada, also has good enough water to support a thriving bass population. Again, the presence of limestone wall rock probably contributes to this situation.

In a few cases, initially acidic lakes have been neutralized by natural processes. Campbell and Lind investigated five coal strip-mine lakes in Missouri. Three of the lakes were still acidic forty years after the end of mining, while the other two had attained neutral pH just fifteen years after mining stopped. The acidic lakes received drainage from coal waste piles and had little vegetation around them. The neutral lakes received drainage from forested areas and farmlands. In the absence of continued acid input from mining waste, the latter two lakes had come to resemble natural lakes, with neutral water and mildly eutrophic conditions. The authors suggested that, in the absence of continued acid drainage from waste piles, the acidic lakes would also have become neutralized.

Good water quality in a pit lake, then, occurs as a result of one or more of the following conditions: low pyrite availability, high carbonate availability, and/or plentiful inputs of organic matter and inorganic nutrients. Unfortunately, without substantial preventive or remedial work, these conditions are limited to relatively few pit lakes.

Improvement of Pit Lake Water Quality

The conditions that have led to the attainment of good water quality in some pit lakes may point to a way to remediate lakes with poor-quality water. The natural remediation of acidic pit lakes has frequently been due to the activity of dissimilatory sulfate-reducing bacteria (SRB). SRB obtain energy for their growth by coupling the oxidation of organic compounds or hydrogen with the reduction of sulfate to sulfide, thus lowering sulfate concentrations. If enough reduced iron is present, sulfide will precipitate. The initially formed iron sulfide is usually amorphous FeS, pyrrhotite or mackinawite (approximate compositions FeS), or greigite (Fe₂S₃). These sulfides are metastable and eventually convert to one of the disulfides, pyrite or marcasite (FeS₂).

The best-known and most widespread sulfate-reducing genera are Desulfovibrio, Desulfotomaculum, and Desulfomonas. Bacteria of these genera are anaerobic heterotrophs, most of which can only utilize a relative handful of organic compounds: lactate, pyruvate, fumarate, malate, acetate, and ethanol. However, other SRB have been found which can oxidize a variety of organics. One Archaeglobus species uses molecular hydrogen to reduce sulfate.

In 1969 Tuttle et al. suggested remediating bodies of water affected by acidic mine drainage by promoting the activity of SRB. This was successfully demonstrated in field experiments on a small stream in Ohio the same year. In laboratory-scale studies, neutralization of acidic pit lake water by SRB was demonstrated on a laboratory scale by Decker and King . Numerous other laboratory studies have demonstrated the application of SRB to acidic mine waters.

Sulfate reduction has the added benefit of precipitating metals and metalloids by reactions of the type

SO₄^2- + M²⁺ + 2 C (org.) ¨ MSø + 2 CO₂

where M is a dissolved metal and C (org) is organic carbon. Moreover, gradual conversion of amorphous iron sulfides to pyrite consumes acid:

2 FeS + ¸ O₂ + 2 H⁺ ¨ FeS₂ + Fe²⁺ +H₂O

The result of these processes is water of near-neutral pH with low levels of dissolved metals and sulfate.

Besides removing acid and soluble sulfates, SRB remove metals from solution by sulfide precipitation. Most transition metals form insoluble sulfides at neutral to mildly acidic pH. Even manganese, which is found in most ore deposits as a carbonate or oxide, can be precipitated as a sulfide under neutral conditions. If conditions suitable for SRB can be established in a pit lake — anoxic conditions and the presence of suitable organic and inorganic nutrients — then it is possible that good water quality may be attained on a time scale of a few months to a few years.

Some workers have advocated simple neutralization of acid lake waters by addition of bases. For example, Rosso reported successful remediation of five coal pit lakes in Kentucky by application of agricultural limestone to the acid-generating areas (i.e., spoil piles) within the watershed of each lake, to the lakes themselves, or both. Some grading and planting were also done in the watershed areas. The lakes’ pH values before remediation were between 3.3 and 4.3. Within eighteen months all had risen to pH 6.7 or higher. It should be pointed out that none of these lakes contained high concentrations of heavy metals other than iron, and only one lake contained high dissolved iron. Sulfate levels before or after the remediation work were not reported. After neutralization the lakes were stocked with channel catfish and largemouth bass, which maintained healthy populations, and also supported large populations of turtles, amphibians, and aquatic invertebrates. Fischer and Guderitz have advocated the application of lime, caustic soda, and other alkaline materials to the waters and drainages of lignite coal pit lakes in Germany to adjust the pH and precipitate iron. The authors did not address the problem of high sulfate levels in the water.

In contrast to simple base addition, biological reduction of acidic water has the potential to remove sulfate and most transition metals. The requirements for bioremediation include the establishment of anoxic conditions in the lake water; reduction of iron (III) and other oxidizing species; establishment of a pH regime suitable for SRB; and provision of the necessary nutrients for SRB.

Anoxic conditions in a pit lake can most easily be established by the addition of organics. A substantial amount of organic matter must be added in order to consume the dissolved oxygen and other oxidizing species, on the order of hundreds of grams carbon per ton of lake water. Therefore, for economic feasibility the added organic matter must be (1) cheap, and (2) locally available. Wood sawdust, spent mushroom compost, hay and straw, partially treated cattle manure, sewage sludge, and waste potato skins are among the organic waste materials that have been more or less successfully used in acid mine water remediation.

When dissolved oxygen in water is depleted, anaerobic bacteria begin using other electron acceptors. These are used in order of their electrochemical reduction potentials, from highest to lowest. Manganese (IV) is used first, then nitrate, nitrite, and iron (III), and finally sulfate . Other redox-active elements which may be present (arsenic, selenium, chromium, etc.) are reduced after species that lie above them in the electrochemical series and before the ones that lie below them. This rule assumes thermodynamic equilibrium, so in cases where kinetics are slow there may be departures from the strict reaction order, but large departures from the rule will be rare.

Several of the reduction reactions consume acid:

2 MnO₂ + C_org + 4 H⁺ ¨ CO₂ + 2 Mn²⁺ + 2 H₂O

2 NO₂^- + 3 C_org + 2 H₂O + 4 H⁺ ¨ 3 CO₂ + 2 NH₄⁺

4 FeOOH + C_org + 8 H⁺ ¨ CO₂ + 4 Fe²⁺ + 4 H₂O

where C_org is organic carbon. Because of iron’s high concentration in most rocks and sediments, iron reduction generally removes the most acid.

When most of the more oxidizing species are consumed, at a redox potential of –75 to –200 mV, SRB are able to begin reducing sulfate, if conditions are otherwise suitable for their activity. Most SRB are not active at pH values less than 5.5, although some sulfate reduction has been reported at pH 4.2. Some dissolved metals may retard the action of SRB. Ueki et al. (1991) reported that the activity of SRB in mine drainage water was almost completely blocked by 1 mM (~50-60 ppm) concentrations of nickel, copper, cadmium, mercury, or zinc. Manganese at the same concentration did not retard SRB activity. However, there are probably some more metal-tolerant strains of SRB around.

Besides promoting reduction of dissolved oxygen and other oxidizers, the added organic waste must also, directly or indirectly, supply the nutrients required by the SRB. This is usually accomplished indirectly by other microorganisms that metabolize the waste matter and excrete the needed nutrients. As with all organisms, enough nitrogen, phosphorus, and trace elements must be present in the water or supplied by the added waste.

Since the water in coal pit lakes tends to be dominated by iron and sulfate, low pH and lack of organic matter are the main obstacles to establishing SRB activity. In metal-mine pit lakes, however, transition metals may be present in high concentrations and so limit SRB activity. It may be necessary to reduce metal concentrations by other means before the SRB will begin reducing sulfate. One possible approach would be to adjust the pH to Ò 6 by adding limestone or another base. This would precipitate many of the transition metals as hydroxides or oxides and simultaneously supply optimum pH conditions for SRB. There would then be a strong probability that addition of organic waste would start sulfate reduction.

An Ounce of Prevention

Those planning open-pit mining activities can work to prevent the development of acid sulfate conditions in pit lakes if they take into account the conditions that lead to acid generation in pit lakes and those conditions that promote sulfate reduction. Preventive measures are usually much less expensive than remediation.

Grading of the site and covering up of waste rock, tailings, and other sulfidic wastes is a good first step, since much of the acid in pit lakes originates as runoff from waste piles. At many metal mines located in arid or semiarid areas, there will be little chance of establishing a vegetative ground cover, but preventing water from running over and through wastes may go a long way toward slowing or stopping the formation of acids. Careful diversion of drainage away from waste rock, isolation of reactive waste in the interiors of waste piles, and covering of waste dumps with impermeable clay caps are common approaches that would substantially decrease metals and acid loading to pit lakes.

An approach that met with a surprising measure of success at one mine was to fill the pit with water as quickly as possible after mining stopped. This approach acknowledged that most oxidation of host rock takes place while the rock is above the zone of saturation. Once the rock is below the water table, the availability of oxygen is reduced and the rate of oxidation is very slow. When the Enterprise Pit, a gold mine in Australia’s Northern territory, was closed in 1992, a major stream was partially diverted into the pit, making the Enterprise Pit a flow-through lake along the stream’s course. After the first wet season the lake was half full, and late in the second wet season it was about two thirds full. The pH of the lake water was 7.2. The lake is expected to serve as an aquatic habitat and as a water resource for the Pine Creek region.

The apparent success in maintaining good water quality at the Enterprise Pit is consistent with findings in other areas. Pyrite-rich mine tailings have been disposed of since 1979 in Anderson Lake, a shallow (8 m) eutrophic lake in Manitoba. Water quality is fairly poor because of process water that is also discharged into the lake, but the water is at pH 7.1 ± 0.3, and there is no evidence of metal or sulfate release from the tailings. The underwater tailings in fact are acting as a sink for dissolved metals. The large amount of organic matter on the lake bottom — a pre-existing condition in the lake — keeps the benthic oxygen demand high and prevents oxidation of the tailings. Mine tailings have also been dumped in suboxic to anoxic sections of several coastal fjords in British Columbia without major metal releases into the water.

The apparent stability of sulfidic waste in anoxic waters supports the idea that fast filling of a pit lake may be one of the best ways to prevent the formation of acid sulfate conditions. It also suggests that the best place to dispose of tailings piles around a pit may be the lake bottom. Advantage can be taken of the tendency of pit lakes to stratify due to their high relative depths. Once sulfidic tailings are in the anoxic hypolimnion of a stratified lake, it is highly unlikely that they will become oxidized.

In many dry areas, for example the western United States, there is not enough water available to fill a pit quickly, and so filling must inevitably take several decades. There may be no way to prevent these lakes from becoming acidic or from containing high metal loadings unless major efforts are taken up front. It would still be desirable to prevent the formation of an acid lake rather than to remediate it after the fact. Removal or covering of tailings and waste rock piles has been mentioned before as a useful preventive measure. Addition of limestone or other alkaline material to the lake to maintain circumneutral pH during filling would slow the oxidation of wall rock, which is catalyzed by acidophilic bacteria below pH 4. Addition of organic waste, where available, would also slow the oxidation of wall rock and tailings and hasten the formation of an anoxic deep layer in the lake. By a combination of the above actions — isolation of waste materials, optimization of drainage, and judicious addition of carbonate rock and organic waste — it might be possible to avoid the environmental impact of an acidic, metal-contaminated lake and the considerable monetary expense of its remediation.

Crater Lakes

These are lakes that form in old volcanic craters. They may or may not have outlets. They have a lot in common with pit lakes physically; their relative depths are high compared to those of most natural lakes. However, because of their locations (commonly on mountaintops), they do not receive as much groundwater input as pit lakes. Hot springs are common on the bottoms of crater lakes.

Crater Lake in Oregon is a lake that formed around 5500 BC after Mount Mazama's last eruption. It is filled by rainfall and snowfall, and its level is maintained by a balance between evaporation and precipitation. There is enough input from hot springs at the bottom, and the spring water is fresh enough, so that the lake is kept mixed by convection.

The overhead shows the cross-section of a small crater lake in the Massif Central in France. Lac Pavin is roughly circular in shape. Its surface area is 440,000 m², and it is 92 m deep; thus its relative depth is 12.3%. This relative depth is greater than those of most natural lakes but smaller than those of most pit lakes. Not surprisingly, the lake is permanently stratified.

The next overhead shows a cross-section of the lake from late autumn. Note that the lake has two thermoclines. There is a warm surface layer, then a cooler layer, and finally a warmer bottom layer. The upper two layers normally undergo turnover once a year, while the bottom layer stays unmixed. The bottom layer is warmer because of hot spring inputs, and it remains unmixed because of its greater density due to its high dissolved solids content.

Here are two more terms for layers in a meromictic lake: The upper layer, which turns over from time to time, is the mixolimnion. The permanently stratified hypolimnion is called a monimolimnion.

The cross-section diagram shows that the monimolimnion is anoxic; the oxycline is more gradual than usual because of the oxygen depletion in the middle layer caused by late-summer stratification. The pH decreases with depth, then increases slightly in the lower part of the monimolimnion, probably because of pyrite formation. The optical transmissivity drops to near zero at the bottom of the oxycline (which coincides with the lower thermocline). At this depth, dissolved Fe²⁺ and Mn²⁺ diffusing across the oxycline react with oxygen and form an opaque layer of solid oxyhydroxides. Thus the monimolimnion is dark as well as anoxic.

Calculation of Activity Coefficients

Activity coefficients afford us a way to allow for nonideal behavior of chemical species in solution.

When ionic solutes are dissolved in water, they increase the charge density in the water and make the water in a sense "more ionic." When the water contains more electrical charge, ionic species become more soluble in the water. At the same time, neutral species, e.g., sugars, dissolved gases like oxygen, and weak acids, become less soluble in highly charged water.

The way we deal with this mathematically is by use of activity coefficients. The activity coefficient of a solute, g _i, is defined as the ratio of the activity of the solute to its concentration:

a_i = g _i c_i

where a is activity and c is concentration.

In general the activity coefficients of ionic solutes are less than one. (See overhead.) Thus, judging by the solute's activity there appears to be a lower concentration of it in solution than there actually is. This means that the solute is more soluble than it otherwise would be. For example, suppose a salt MX has a solubility product constant K.

K = {M⁺} {X^–}

The solubility of MX is the concentration of either M⁺ or X^– when MX is in saturated solution. Therefore, if MX is the only dissolved substance, [M⁺] = [X^–], and the solubility of MX can be written as the square root of the product ([M⁺] [X^–]). In an ideal solution, the activity coefficients are unity, {M⁺} = [M⁺], {X^–} = [X^–], and the solubility of MX is equal to (K)^1/2.

Now let's look at a real-world case rather than an ideal case. Suppose the activity coefficient of M⁺ is 0.85 and that of X^– is 0.90. Then

K = {M⁺} {X^–} = 0.85 [M⁺] « 0.90 [X^–]

K = 0.765 [M⁺] [X^–]

In other words, MX is 14% more soluble than it would be in an ideal solution where activity coefficients are all equal to one.

In a similar way, the activity coefficients of neutral molecules become greater than one in solutions with high ion concentrations, and the solubilities of those neutral substances become smaller. This is the cause of the phenomenon called "salting-out" whereby neutral or nonpolar substances precipitate from solution when salt is added.

In the case of a dilute solution we are often able to disregard the changes in ion activity with total electrolyte concentration in water; the activity coefficients remain very close to unity in solutions around a few millimolar concentration or less. However, water in a pit lake commonly contains high levels of dissolved matter, and activity corrections are necessary if we wish to calculate things like saturation indices, degrees of acid dissociation, etc.

In a matrix like Berkeley Pit water, it is necessary to know concentrations in order to calculate activities. Moreover, some concentrations, such as those of ionized and unionized sulfuric acid, depend on activities, so we also need to know activities in order to calculate concentrations. This circular dependence makes it necessary to carry the calculations through multiple iterations until we reach a self-consistent set of activities and concentrations.

The total ionic strength, I, of a solution, is defined as

where c_i is the molal concentration of the i^th ionic species, z_i is its charge, and there are n ionic species in all.

In general, the activity coefficient of a species in solution is a function of the solution's ionic strength. The simplest equation to calculate activity products is the Debye-HŸckel limiting law, which is applicable for solutions with ionic strength under 5« 10^–3 M:

The extended Debye-HŸckel equation applies to solutions with ionic strength of 0.1 M or less:

A and B are parameters obtained by the formulas

A = 1824948 r ₀^1/2 (e T)^–3/2

and

B = 50.3 (e T)^–1/2

where T is the absolute temperature (K), r ₀ is the density of water (g cm^–3), and e is water's dielectric constant. At 15¡ C, A = 0.500787 and B = 0.3269. The ion-size parameter, a_i, is determined separately for each ion.

The Truesdell-Jones modification of the extended Debye-HŸckel equation is necessary in concentrated solutions like those in acidic pit lakes:

The empirical parameter, b, is determined experimentally for each dissolved species. The following table lists values of a_i and b for some common ionic species.

Ionic strength is strongly influenced by the degree of ionization of sulfuric acid, and the acid's second ionization is a strong function of ionic strength. Another equilibrium with a strong influence on ionic strength is the formation of the neutral calcium sulfate complex, which is governed by the equilibrium

This equilibrium interacts with gypsum's solubility product:

{Ca²⁺}{SO₄^2–} = 2.45« 10^–5

Both of these equilibria are also functions of ionic strength.

There are computer programs that can perform these calculations. It is important, however, to understand the theory behind these calculations, and to be sure an appropriate formula is being used to calculate the activity coefficients for a given set of solution conditions. The Truesdell-Jones equation is good up to about 1 M ionic strength. Above that level (e.g., for seawater, brines, or some hot-spring waters) an even more elaborate calculation is needed.

Examples

1. Calculate the activity coefficients for a 1 mM solution of sodium chloride.

Ionic strength:

I = ¸ [0.001 (+1)² + 0.001 (–1)²] = ¸ [0.001+0.001]

I = 0.001

At this ionic strength, the simple Debye-HŸckel equation is adequate.

Activity coefficient for Na⁺:

log g = –0.5 (+1)² [0.001]^1/2

log g = – 0.5 (1) (0.03162) = – 0.0158

g = 0.964

Activity coefficient for Cl^–:

log g = – 0.5 (–1)² [0.001]^1/2

g = 0.964

2. Calculate the activity coefficients for a 0.2 M solution of sodium sulfate.

Concentrations are 0.4 M Na⁺ and 0.2 M SO₄^–. Ionic strength is

I = ¸ [0.4 (+1)² + 0.2 (–2)²] = 0.5 [0.4+0.8]

I = 0.6

At this ionic strength, the Truesdell-Jones modification must be used. We need to specify a temperature, since the density of water and its dielectric constant are functions of temperature. If we assume T = 25¡ C, then A = 0.5085 and B = 0.3281.

Using the values of a_i and b listed in the table, and noting that the factors of 10⁸ in B and 10^–8 in a_i cancel,

For sodium:

log g = –0.1063

g = 0.783

For sulfate:

log g = –0.7176

g = 0.192

To previous lecture: Lecture No. 20. Pit Lakes

To next lecture: Lecture No 22. Pit Lakes III

Back to list of lecture notes