Lecture No. 23. Special Topics I

Natural High Arsenic Systems

We have seen examples of environments contaminated with arsenic by mining or other industrial activities. There are numerous places in the world where high, often hazardous, arsenic levels exist naturally. We will look at a few of those environments.

I. Mono Lake, California

Mono Lake is a highly saline lake east of the Sierra Nevada and more or less due east of Yosemite National Park. The lake's salinity is about 90‰, which is about 2.5 times that of seawater. The salinity used to be about half of that value, around 45‰, but that was before the Metropolitan Water District (MWD) began diverting freshwater streams away from the lake and into the Los Angeles Aqueduct. A couple of decades ago the lake's level had fallen enough to open up a dry corridor to some islands in the lake that were seagull rookeries, and coyotes and other predators began robbing the nests. This led to a Federal order to bring the lake level back up far enough to isolate the island again, and the water withdrawals have since been a little reduced.

Mono Lake is in an active volcanic region; the Mono Craters last erupted around A.D. 1330. The local rock is high in arsenic. The lake itself, thanks in part to mineral springs in its floor, is very high in arsenic. Dissolved arsenic in lake water runs as high as 15 mg/L at 20 m depth.

The lake usually overturns once a year. When it overturns, As (III) from its hypolimnion is brought to the surface, so that arsenic is regularly recycled through the entire lake volume. The newly surfaced As (III) is oxidized to As (V) by dissolved oxygen. Much of this arsenate binds to iron oxyhydroxides in the water.

There is a day-night periodic cycle of dissolved arsenic. During the day, sunlight promotes reduction of iron (III) to iron (II) by the reaction

Fe(OH)₂⁺ + hn ¨ FeOH⁺ + OH

where OH is a neutral free radical. The reduction of iron liberates some arsenate that had been bound to the iron, and dissolved As concentrations increase. During hours of darkness the arsenic is sorbed again to iron (III) oxyhydroxides, and the dissolved As concentration decreases.

II. Long Valley Caldera, California

A short distance south of Mono Lake is the Long Valley Caldera. This is a rhyolite-type resurgent caldera that has not erupted in a long time. When it does erupt, it is in the same class as the Yellowstone Volcano. Its last eruption was around 700,000 years ago. Some vulcanologists believe it is overdue for a new eruption. As is usual with silicic lavas, the Long Valley lava is high in arsenic. Hot springs in the Long Valley area produce water that is very high in arsenic. The water in Hot Creek is fed by a large number of these springs before it joins the Owens River. Since the Los Angeles Aqueduct takes the entire flow of the Owens River, all that arsenic ends up in the MWD's water supply. It is an indication of the high arsenic concentration in the hot springs that Hot Creek downstream of the springs contains about 200 m g As/L, and the entire content of the Los Angeles Aqueduct actually violates drinking water arsenic standards some of the time.

Janet Hering of Caltech gave a seminar on this subject earlier this semester.

III. The Waikato River, New Zealand

The Waikato River on North Island is the longest river in New Zealand. It picks up a fair amount of arsenic from hot springs near Lake Taupo and this is aggravated by water discharges from the Waikarei power station, which taps high-arsenic geothermal water. There are a number of hydro reservoirs on the river. One of these is Lake Okahuri, which receives the water discharged from the Waikarei power plant. The lake is stratified for part of the year. During that period the surface water contains about 50 m g As/L, and the hypolimnion contains 150-250 m g/L. As in the case of Mono Lake, the As (III) is carried upward during lake overturn and oxidized to As (V). Since the lake is a power-generating reservoir, hypolimnetic water is drawn through the penstocks to the generators and discharged downstream, carrying its burden of As (III).

IV. The Madison and Missouri Rivers, Montana

The Madison River, which forms from the confluence of the Gibbon and Firehole Rivers, drains the arsenic-rich western side of Yellowstone Park. In the park, the Madison carries about 250-370 m g As/L in solution. In the Madison Valley outside the park, surface and ground waters both carry elevated levels of arsenic. In the upper part of the valley, surface and ground water range between 16 m g/L and 176 m g/L, while in the lower valley the arsenic content in groundwater is 25-50 m g/L. For a time it was thought that irrigation with Madison River water was the cause of the high arsenic in groundwater. However, further study by David Nimick showed that the main source of recharge to the aquifer was the river itself. In addition, the entire aquifer consists of Pleistocene-age alluvial deposits from the Madison, which may be expected to contain high arsenic levels. If anything, the aquifer materials could be expected to release arsenic into the water rather than act as a sink.

The dissolved arsenic in the river carries over into the Missouri. The first overhead shows a log-log plot of stream discharge vs. dissolved arsenic. Most of the numbers follow the Madison River dilution line pretty closely, until the river reaches the Missouri Breaks area at Virgelle and Landusky. In that area the river receives a large input of clean sediment from the Marias, Teton, and Judith Rivers. The sediment in the Madison River is already loaded with as much arsenic as it can carry, so the Missouri between Three Forks and the Missouri Breaks carries practically all of its arsenic conservatively in solution. The drop in dissolved concentration at Virgelle and Landusky is due to sorption onto clean sediments from the tributaries. The dissolved arsenic largely stays below the Madison River dilution line after that; note the low arsenic numbers below Fort Peck Lake.

Also note from the overhead that the Madison River water at Ennis exceeds the MCL for drinking water, and it frequently exceeds this limit at Toston (near Townsend). The arsenic level in the river is high enough so that Helena and Great Falls, which draw their water supplies from the river, treat the water to remove arsenic.

It should be noted that the Missouri River at Three Forks is predominantly Madison River water during the summer, because there are heavy withdrawals of irrigation water from the Gallatin and Jefferson Rivers. During the winter each fork contributes about 1/3 of the total discharge.

The next overhead is a plot of total recoverable arsenic vs. dissolved arsenic in the river. Note that the total recoverable arsenic is nearly the same as total arsenic except in the samples from Virgelle and Landusky. This again indicates that some of the dissolved arsenic is moving into the suspended phase in this area. The samples from the Missouri below Fort Peck Lake show arsenic to be again almost completely dissolved. The water in Fort Peck Lake has a residence time of several years, and the lake is an effective sediment trap. Therefore the river water immediately below the lake is low in sediment-bound arsenic (because it is low in sediment), and almost all the arsenic is dissolved.

The third overhead shows daily periodic variation of pH and dissolved arsenic at various points along the river. At West Yellowstone the pH swings through a range of about 0.8 between day and night, but there is little variation in dissolved arsenic. Farther downstream, at Norris, Three Forks, and Virgelle, the dissolved arsenic tracks pH with a slight phase delay. The numbers are subject to too much variation at Toston to say for sure what is happening there. The sampling site is not too far downstream of Toston Reservoir, a shallow lake that probably has a homogenizing effect on the river and smoothes out periodic variations. Note that the periodic arsenic changes are real, but that they are small and have little effect on overall arsenic levels. Also note that the pH variation at Virgelle is smaller than upstream; the river is becoming better buffered as it gains bicarbonate.

The dominant dissolved ions in the Madison River within Yellowstone Park are Na⁺, Cl^–, and HCO₃^–. In the lower Madison Valley the mix has changed to Ca²⁺, Na⁺, and HCO₃^–. At Toston the river has picked up some magnesium, so that the dominant ions are Ca²⁺, Mg²⁺, Na⁺, and HCO₃^–. This situation persists until Virgelle. During the time of year when there is heavy runoff from the shales of the Judith Basin, the water is dominated by Na⁺ and SO₄^2–; during the rest of the year, Ca²⁺, Mg²⁺, and HCO₃^– dominate. Below Fort Peck the river has important contributions from both sodium sulfate and calcium-magnesium bicarbonate sources.

Fort Peck Lake, and, to a lesser extent, Canyon Ferry Reservoir may be acting as important arsenic sinks. All the other lakes and reservoirs along the river are too shallow to act as effective traps.

V. High-Arsenic Groundwater

High-arsenic groundwater has constituted a health problem in several areas around the world.

• Artesian wells were drilled into a deep aquifer in Taiwan and served from 1961 to 1985 as a drinking water source for about 100,000 persons. The water contained between 100 and 1800 m g As/L. The people drinking the water had a high incidence of arsenic-related ailments including cancer and a gangrenous condition called "blackfoot disease" that entailed rotting and blackening of the soles of the feet.
• About 200,000 persons in Regi—n Lagunera in Durango, Mexico, used well water containing about 400 m g/L arsenic from 1963 to 1983. About 20% of the population showed symptoms of chronic arsenic poisoning.
• High arsenic levels in water around Monte Quemano, C—rboba Province, Argentina were the apparent cause of high cancer rates among the residents.
• From 1959 to 1970 about 100,000 persons in Antofagasta, Chile, used drinking water from a source containing about 800 m g/L arsenic. At least 12% of them had arsenic-related skin problems ranging from hyperkeratosis to cancer.
• The groundwater in an area of Millard County, Utah that was consumed by about 250 persons contained 180-210 m g/L arsenic. The residents had a slightly elevated incidence of arsenic-related abnormalities. On the other hand, residents of areas in Lane County, Oregon, Lassen County, California, and North Star Borough, Alaska, who consumed water containing over 50 m g/L arsenic have shown no apparent arsenic-related health problems and no increased incidence of cancer.
• However, by far the worst arsenic-related public health disaster has been in Bengal.

Bengal consists of the Ganges-Brahmaputra-Jamuna delta. It is divided (along religious lines) between the Indian State of West Bengal and the country Bangladesh. Although the area has a very wet climate (about 200 cm/y of rainfall), most of the rain falls during the summer monsoon. The water supply has long been a problem there. That is to say there is plenty of surface water, but there is very little that is not contaminated by cholera, typhoid, and various other diseases. The World Health Organization and the Indian government promoted a program of installing tube wells as an alternative rural water supply. The wells are dug with hand augers, are pumped by hand, and can deliver about 2 L/min.

The well water is free of bacteria and parasites. Unfortunately it is not free of arsenic. The overhead shows a map of West Bengal and its location within India. The stippled parts of the map indicate six districts of West Bengal that have seriously elevated arsenic in the groundwater. (Calcutta is within the westward bend of the Hugli River just above its mouth.)

The well water in the affected districts contains several hundred micrograms arsenic per liter in most cases. The source of the arsenic appears to be pyrite in two formations that serve as aquifers. The groundwater has been seriously drawn down during the past several decades to support irrigation during the dry (winter) monsoon. This has exposed the pyrite to oxic conditions and it has oxidized. Since the pyrite has a high arsenic content, the arsenic is released into the groundwater in soluble form. Thus the arsenic source is natural, although it has been aggravated by irrigation.

The population of the affected districts is about 30,000,000. The populations of the villages using arsenic-contaminated water number about 800,000, of whom 175,000 show symptoms of arsenic poisoning. The mildest symptoms are skin discoloration and mild keratosis. In the more severe cases, the symptoms include hyperkeratosis, blackfoot disease, and cancer. To make matters worse, the arsenic poisoning symptoms resemble those of leprosy, so that people suffering from poisoning are often shunned by friends and family. Some wives have been sent back to their parents when they showed the symptoms.

The same thing is happening in adjacent parts of Bangladesh. There is also reason to suspect that the parts of West Bengal and Bangladesh that do not now have high arsenic levels in their groundwater will eventually have high arsenic as the drawdown continues and the pyrite oxidation spreads. Eventually several tens of millions of persons may be exposed to arsenic poisoning.

Public agencies have done little so far to correct the situation. Villagers who originally listened to public health authorities who told them to stop drinking river water and dig wells may be reluctant to listen to the same authorities who now are telling them to stop using the wells. Besides, drinking river water often kills people quickly, by disease, rather than gradually by arsenic poisoning.

The only thing that will stop the poisoning in the long run and stave off a disaster of epic proportions (tens of millions of deaths) would be sewage treatment plants all over India, Bangladesh, and Nepal or an extensive system of reservoirs and water works in Bengal to purify surface water. No one knows where the money to build such a system would come from.

All of the cases of chronic arsenic poisoning seen so far have resulted from arsenic levels over 100 m g/L, and usually several hundred m g/L. The current drinking water MCL of 50 m g/L may or not be completely safe, and the EPA is considering dropping the MCL to somewhere in the range 2-20 m g/L.

Acid Sulfate Soils

These are soils with their chemistry dominated by the acid products of pyrite oxidation. Some of them are the results of mining practices, which we have covered in some detail already. There are several other sources of the acid sulfate condition, including reclamation of land submerged under the sea, glacial processes, heavy construction, and even severe acid rain.

• Seawater is high in sulfate (0.26%). Seabottom sediments are anoxic below the top few millimeters, except where bioturbation keeps them oxic. In the anoxic sediments there is a great deal of pyrite. Land reclaimed from the sea for farming, such as tidal flats, turns very acidic as soon as it is drained. In some areas, such as the Netherlands, carbonate bedrock neutralizes the acid and the soil soon becomes productive. In other areas, such as Indonesia and Malaysia, the bedrock has little or no acid-neutralizing capacity, and a great deal of limestone has to be added.
• In southern Ontario the Pleistocene glaciations exposed metamorphic basement rocks in many locations. This rock contains large amounts of pyrite, and the soil that develops from this rock is acidic. The acidic soils in southern Ontario are characterized by many yellow, brown, and white flecks consisting respectively of jarosite, goethite, and gypsum. Farmers have to use periodic applications of lime to keep the soil productive.
• Deep excavations can expose sulfidic bedrock in the same way that mining does. For example, construction of the Divide section of the Tennessee-Tombigbee Canal in northern Alabama exposed pyritic rock, which became the source of acidic drainage into the canal and from there into the Tennessee and Alabama Rivers.
• Acid rain is the process of wet and dry deposition of acid onto the land. Sulfuric and nitric acids are the two main components.

What all of these acidic soils have in common is low pH and high levels of sulfate; therefore they are collectively called acid sulfate soils. Worldwide there are estimated to be 12 million hectares of acid sulfate soils.

The typical mineral assemblage in acid sulfate soils is listed in the following table:

The fact that large amounts of hydrated iron oxides coexist with acidic sulfates suggests that the soil minerals are not in equilibrium.

Exactly what other minerals may be present in the soils depends on both the degree of acidification and the parent materials. In southern Ontario, where the soil acidity is due to the weathering of igneous and metamorphic rocks, the soil contains high levels of swelling clays formed by acid reacting with micas. Near Nieukoop, Netherlands, the former acid sulfate soils are dominated by illite, quartz, feldspars, calcite, and dolomite. The acid sulfate soil in Pulau Petak, Indonesia, was dominated by kaolinite, illite, smectite, and quartz. The general pattern is one of micas being converted to swelling clays, calcite being converted to gypsum, and iron minerals being converted to jarosite and iron oxides. In the lowest-pH soils, substantial amounts of iron (III) and aluminum are in solution as hydroxy cations, and iron and aluminum phytotoxicity becomes a problem.

Mine Soils

In the Appalachians, blasted overburden from coal mines is frequently placed on the surface to be reclaimed. This material, which may or may not contain significant amounts of sulfides, is exposed to the usual soil-forming processes controlled by climate, biota, and time. After a few years of weathering, this material is referred to as a mine soil. This soil tends to be rocky, infertile, and frequently acidic. However, the original pre-mining soil on the sites was also frequently thin, rocky, acidic, and infertile.

In the western U.S., around open-pit metal mines, mine soils tend not to develop, or to develop very slowly, because of the generally arid conditions around the mines and because the waste rock frequently consists of igneous rock rather than easily-weathered sandstone and shale.

The suitability of a mine soil as a substrate for vegetation depends on its mineralogy and petrology, its physical state, its acid content, its acid-generating potential, and the kinds of amendments added to it.

When we consider metal-sulfide deposits, we find that some types produce much more acid than others. The worst by far are the massive sulfide deposits, which can produce drainage water with negative pH, i.e., with hydrogen ion activities greater than 1 M and sometimes greater than 10 M. At the other end of the spectrum are the quartz-adularia and adularia-sericite deposits, which have very low acid-generating potential.

Another important determinant of soil acidity is the acid-buffering capacity of gangue minerals and host rock. Some soils with high carbonate contents remain neutral or alkaline. However, these soils can have their own problems. Reaction of calcite with sulfuric acid produces gypsum, and the result, especially in dry climates, is a gypsum caliche layer.

Approaches to Remediation of Acid Soils

In a few cases, such as those of coal mines or polders on limestone or dolomite formations in humid climates, the soil will remain neutral, and little or no action needs to be taken. However, if metals other than iron are involved, even a neutral soil may contain toxic levels of those metals.

In most cases, some remedial action must be taken. Those actions may take any of several forms.

1. Deep Liming of the Soil

Liming — addition of calcium carbonate or hydroxide — has long been advocated as a way to neutralize acids in soils. Usually this entails adding lime to the top 10-30 cm of the soil — roughly to plow depth — and it has little effect on mine soil fertility. The deeper soil remains acidic, it will continue to produce more acid. Plants will thrive until they put down roots past the neutralized zone, at which point they may stop growing or may die. Recently, however, a study was done in which lime was added to a depth of 120 cm. In this case the soil was rendered suitable for plant growth, and the improvement appeared to be permanent.

The test site was an abandoned coal waste site near Thurber, Texas. A number of conventional lime application methods were tried on the site as well as some experimental methods. Most of the methods tried did not alleviate the acidity problem below the top few centimeters of soil. One method used by a group from Montana State University, however, seems to have worked. A large moldboard plow (see overheads) with 1.2 m (4 ft) penetration depth was used, along with a spray system that delivered calcium hydroxide slurry evenly through the entire depth. A D-9 Caterpillar tractor was needed to pull the plow, and a full-size tank truck was needed to supply the slurry. The only economical way to deliver the Ca(OH)₂ slurry was to mix water with calcium oxide (quicklime) in the field. The resulting slurry was therefore very hot.

The MSU group tried the deep lime incorporation method at a second coal mine site at Rockdale, Texas. The soil pH was successfully raised above 5.5 (in most cases above 6.5), and it stayed above that level after 6 months. They also tried the method without success in the Silver Bow Creek tailings. (Their plow depth ran below the water table in the latter case, and the level of acidity in the tailings may also have been very high.)

The deep lime incorporation method has promise in treating acidic soils. It does not come cheap, however. The estimated treatment cost is about $2600/ac, including the cost of both equipment and chemicals. This cost might dissuade some potential users. As was mentioned above, the method cannot be used if the water table is less than 1.2 m deep. Finally, in some cases liming can worsen a soil's nutritional deficiencies. In Australia, adding lime to soil at the site of a lead-zinc mine raised the pH and lowered zinc levels below phytotoxic thresholds but did not improve the overall situation. The higher pH rendered soil boron unavailable to plants. Liming is also known to decrease phosphorus availability to plants. In some cases, fertilizing with selected nutrients may solve this problem. In other cases, any pH that immobilizes contaminant metals also may immobilize necessary plant nutrients.

2. Addition of Organic Matter

Organic amendments have been used on mine soils for several decades. Amendments have included sewage sludge, paper-mill sludge, compost, and mixtures of sludge and wood chips. The most successful treatments have been the sewage sludge and the sewage sludge mixtures. As a result, sludge has become a big business in the Appalachian mining belt. Since the early 1980's Philadelphia has been marketing a mixture of cake sludge, wood chips, and composted sludge called "Philadelphia Mine Mix."

However, concern has been raised about the high levels of heavy metals in urban sewage sludge. Numerous workers have reported that soil repeatedly treated with sewage sludge picked up toxic levels of heavy metals. For example, Benninger-Truax and Taylor reported in 1993 that soil treated regularly with sewage sludge for eleven years had acquired elevated concentrations of cadmium, copper, lead, and zinc, and that all of these metals were present in very bioavailable forms. Several European countries have severe limits on the amount of sludge that can be applied to land.

On the other hand, one-time sludge applications do not cause problems. Sopper reported in 1992 the results of a 12 year study of two mine spoil sites in Pennsylvania that were treated once or twice with sludge mixes and acquired no elevated metals.

The mechanism by which sludge application improved soils is fairly well understood. It adds organic matter to soil that has practically none, and it improves the soil structure. It also adds nutrients usually lacking in mine soils. Finally, and probably most importantly, it promotes the development of a healthy soil microbial community, which is vital for the development of soil nutrient cycles. However, it does not neutralize acid in the soil. In fact, sludge addition tends to drop the pH by promoting nitrogen mineralization and nitrification. Nitrification generates acid by the reaction

2 NH₄⁺ + 3 O₂ ¨ 2 NO₃^– + 4 H⁺

To counteract this acid-producing process, limestone is usually added with sludge.

Other organic matter has been added to soils, but sludge remains the most cost-effective additive because of its plentiful supply and low cost.

3. Isolation of Sulfidic Strata

In cases where there are deep sources of sulfide in contact with oxygenated ground water, an alternate approach has been to concentrate on the top meter or so of soil and take steps to isolate the deeper strata. Researchers at the University of West Virginia have tried applying phosphatic clay slurry between pyritic spoil and the soil layer. The soil layer may be imported topsoil or it may be a relatively low-acidity mine spoil. Phosphatic clay is a byproduct of fertilizer manufacture. Its approximate composition is 25% apatite, 60% smectite clay, and 15% other phosphate minerals. The smectite is a swelling clay and tends to prevent water transport between layers. Moreover, if the sealant layer is penetrated and dissolved iron migrates upward, the phosphate present should precipitate the iron as strengite (FePO₄) and vivianite (Fe(PO₄)₂ á 8 H₂O).

Hanson et al. (1984) suggested the use of a 15 cm layer of glacial till as an impervious layer over mine spoils. Since the till itself is a fair growth medium, no organic amendments would be needed. The till's low permeability might prevent upward migration of acidic groundwater.

To previous lecture: Lecture No 22. Pit Lakes III

To next lecture: Lecture No. 24. Special Topics II

Back to list of lecture notes