Environ. Sci. Technol.

2006, 40, 752-758

Element Flows Associated with

Marine Shore Mine Tailings
Deposits
BERNHARD DOLD*
Centre dAnalyse Minerale, Institute of Mineralogy and
Geochemistry, Batiment Humense, University of Lausanne,
CH-1015 Lausanne, Switzerland

From 1938 until 1975, flotation tailings from the Potrerillos-El

Salvador mining district (porphyry copper deposits)
were discharged into the El Salado valley and transported
in suspension to the sea at Chanaral Bay, Atacama
Desert, northern Chile. Over 220 Mt of tailings, averaging
0.8 ( 0.25 wt % of pyrite, were deposited into the bay, resulting
in over a 1 kilometer seaward displacement of the
shoreline and an estimated 10-15 m thick tailings accumula-
tion covering a 4 km2 surface area. The Chanaral case
was classified by the United Nations Environmental
Programme (UNEP) in 1983 as one of the most serious
cases of marine contamination in the Pacific area. Since
1975, the tailings have been exposed to oxidation, resulting
in a 70-188 cm thick low-pH (2.6-4) oxidation zone at
the top with liberation of divalent metal cations, such as
Cu2+, Ni2+, and Zn2+ (up to 2265 mg/L, 18.1 mg/L, and 20.3 mg/
L, respectively). Evaporation-induced transport capillarity
led to metal enrichment at the tailings surface (e.g. up to 2.4%
Cu) in the form of secondary chlorides and/or sulfates
(dominated by eriochalcite [CuCl2H2O] and halite). These,
mainly water-soluble, secondary minerals were exposed
to eolian transport in the direction of the Village of Chanaral
by the predominant W-SW winds. Two element-flow
directions (toward the tailings surface, via capillarity, and
toward the sea) and two element groups with different
geochemical behaviors (cations such as Cu, Zn, Ni, and
oxyanions such as As and Mo) could be distinguished. It can
be postulated, that the sea is mainly affected by the
following: As, Mo, Cu, and Zn contamination, which were
liberated from the oxidation zone from the tailings and
mobilized through the tidal cycle, and by Cu and Zn from
the subsurface waters flowing in the El Salado valley (up to
19 mg/L and 12 mg/L Zn, respectively), transported as
chloro complexes at neutral pH.
1. Introduction
Marine deposition of sulfide mine tailings is a worldwide
problem (1-4). Locally, it can result in large accumulations
of material that greatly impacts marine life due to metal
release and high sedimentation rates (3, 5-7). The effect on
the marine ecosystem at Chanaral, located in the Atacama
desert, northern Chile (Figure 1), from more than 50 years
of porphyry copper tailings deposition from the Potrerillos-
El Salvador mining district was studied in detail (2, 7-13).
The metal speciation of the sandy beaches in direct contact
* Corresponding author phone: +41 021-692-4327; fax: +41 021-
692-4315; e-mail: Bernhard.Dold@unil.ch.
752 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 3, 2006
FIGURE 1. Location of the sampling points CH1 to CH4 at the Chanaral
shore tailings deposit, Northern Chile. For better visibility, the depth
of the profile is strongly exaggerated.
with the sea was studied by sequential extractions (13). The
Cu speciation between the beach pore water and the marine
water interface was investigated by diffuse gel technique (14).
The study presented herein is the first to investigate the
geochemical processes leading to element flows from the
deposition of mine tailings in a marine shore setting to the
different ecosystems (i.e. seaward and inland, in this case,
toward the Village of Chanaral and surroundings), and the
subsequent bioavailability of the released toxic elements, as
a function of the climatic and the hydrological setting.
1.1. Regional Geology and Ore Geology. The source of
the tailings deposited in the Bay of Chanaral was the El
Salvador-Potrerillos mining district, Chile (Figure 1). From
1926 to 1971, the Andes Copper Mining Company owned
the El Salvador-Potrerillos mining district. Since 1971, the
district has been owned and operated by CODELCO. The
Potrerillos porphyry copper deposit was exploited until 1959.
That same year, El Salvador started production. Therefore,
it can be assumed that the lower part of the tailings deposited
at Chanaral are from Potrerillos, overlain by tailings from El
Salvador. The Potrerillos deposit (1.8 106 t Cu) is located
in the Atacama Desert, 120 km east of Chanaral, and is
characterized by three hypogene alteration and mineraliza-
tion events and an important supergene alteration. The early
potassic and propylitic alterations, are responsible for the
K-feldspar, biotite, chlorite, quartz, ankerite, and anhydrite
content. Associated sulfides are chalcopyrite, bornite, pyrite,
molybdenite, minor enargite, sphalerite, and galena (15). The
El Salvador porphyry copper deposit (5.7 106 t Cu; 41 Ma)
is located about 100 km east of Chanaral (Figure 1). Early
mineralization is characterized by distinctive quartz veins
and largely disseminated K-silicate assemblages of alkali
10.1021/es051475z CCC: $33.50 2006 American Chemical Society
Published on Web 12/30/2005
feldspar-biotite-anhydrite-chalcopyrite-bornite or chal-
copyrite-pyrite. K-silicate alteration converted hornblende
phenocrysts to biotite-anhydrite-rutile, ilmenite to hema-
tite-rutile, and titanite to rutile-anhydrite. Secondary Cu-
sulfides (chalcocite and covellite) extensively replaced chal-
copyrite and bornite, but coated pyrite with little or no
replacement (16, 17).
1.2. Mining and Treatment Processes. Potrerillos and El
Salvador were both underground mining operations. During
flotation, the ore minerals were separated from the gangue
by a collective Cu-Mo sulfide flotation. The flotation process
was conditioned by lime to a pH of 10.5, which was the pH
of the tailings upon discharge.
1.3. History and Geographical Setting of the Tailings at
Chanaral. From 1938 until 1975, most of the flotation tailings
from the Potrerillos-El Salvador mining operations were
discharged and carried in suspension via the El Salado valley
directly to the sea at the Chanaral Bay (from 1938 to 1971,
by Andean Copper Mining Company; from 1971 to 1975, by
CODELCO). Until 1959, the tailings came from Potrerillos,
and after this date, they came from El Salvador. From 1975
to 1989, the tailings were deposited at Punta Palitos, 15 km
north of the Bay of Chanaral. In 1989, a lawsuit by the Village
of Chanaral stopped deposition at Punta Palitos. Thereafter,
the tailings were deposited in the Pampa Austral tailings
impoundment located close to the El Salvador Mine.
1.4. Climate and Hydrological Setting of the Tailings
Deposition Area. Precipitations in the Atacama desert (hyper-
arid) are negligible (30 mm/a), apart from some rainfalls
associated with the El Nino event (18). Thus, the hydro-
logical system of the shore mine tailings deposit at Chanaral
is mainly controlled by the tides. The difference between
high and low tide in the Bay of Chanaral is about one meter.
During high tide, water infiltrated the tailings. During low
tide, this flow retreated toward the sea. The high-energy wave
action of the Pacific Ocean along the Chilean coast produced
a grain size fractionation, resulting in a sandy beach and the
transportation of the fine particles seaward. The episodically
very strong winds in the Chanaral area, from a predominantly
W-SW direction, formed up to 1.5 m high sand dunes, which
migrated on the tailings surface toward the Village of
Chanaral. The tailings surface, which is not covered by the
dunes, show abrasion marks with the same wind direction.
Subsequently, a strong eolian transport of tailings and sand
particles was observed in the direction of Chanaral, resulting
in the need to remove tons of sandy material from the house
yards in Chanaral after each strong wind event, principally
in the airport area (Figure 1).
2. Experimental Section
2.1 Sampling and Field Measurement. In October 2003, three
piezometer nests were installed to a maximum depth of 9 m,
along a profile in the southern part of the tailings deposit,
directly in front of the Village of Chanaral (CH1, CH2, and
CH3; Figure 1). Using percussion soil sampling equipment,
solid samples down to a max 3 m depth, were obtained at
each piezometer site. At point CH4, located at the beach,
about 5 m from the sea, a one meter core was taken. The
Playa Pan de Azucar (PPA) beach, located 30 km north of
Chanaral in the National Park Pan de Azcar, was sampled
as an uncontaminated natural background beach.
2.1.1.Tailings Samples. Tailings samples were collected
using stainless steel tubes, 3 cm in diameter. The upper meter
of the core was continuously sampled at 20 cm intervals. The
deeper sediments (max 3 m depth) were selectively sampled
according to changes in granulometry, mineralogy, and color
(oxidation state). Paste-pH was measured using a WTW pH
meter (WTW pH 323), with a special pH electrode (WTW-
Sentix RP), inserted directly into the moist tailings. The tailings
samples were air-dried and stored for mineralogical and
geochemical study.
2.1.2. Pore-Water Sampling. 2.1.2.1.Vadose Zone Sam-
pling. Pore-water from the vadose zone was extracted from
sealed core samples (up to 2 m depth), collected in 7.63 cm
diameter thin-walled aluminum casings. The cores were cut
into 50 cm-long sections. The cores were instantly frozen
(-20 C) at the field site, and maintained frozen until pore-
water extraction. The cores were thawed at room temperature
(25-30 C). The pore water was then displaced from 25 cm-
long cores using epoxidized soybean oil (PARAPLEX), which
is immiscible with water, as modified after the method
described by Patterson et al. (19). Determination of pore
water pH and Eh were made immediately after pore-water
extraction. The calibration of the pH electrode was performed
using pH 4.0 and 7.0 standard buffer solutions. The perfor-
mance of the Eh electrode was controlled by comparison to
Zobells solution and Lights solution (20, 21). All Eh values
were corrected to the standard hydrogen electrode (SHE).
All samples were filtered through 0.2 m nylon syringe filters
and split into two aliquots. One was acidified with HNO3 to
a pH <2 for cation analysis. The second untreated sample
was frozen for anion analysis.
2.1.2.2. Saturated Zone Sampling. Three piezometer nests,
each with 4 individual drive-point piezometers (Solinst), were
installed at the CH1, CH2, and CH3 to obtain water samples
from the saturated zone. All piezometers were developed by
repeated pumping. Water samples were collected using a
peristaltic pump, and pH, Eh, and temperature were mea-
sured using sealed flow-through cells. All water samples were
filtered (0.2 m; regenerated cellulose) and frozen at -20 C
until analyzed. Samples for cation analysis were acidified to
pH <2 with suprapure HNO3. Alkalinity determinations were
made in the field using a Hach digital titrator.
2.2 Analytical Methods. 2.2.1. Mineralogy and Geochem-
istry of the Solids. Forty-five air-dried tailings samples were
submitted for mineralogical study, using thin and polished
sections microscopy and X-ray diffraction (XRD). All samples
were characterized by XRD using a Philips diffractometer
3020 with CuKR ( ) 1.54056 ) and monochromator at the
Mineralogical and Geochemical Institute, University of
Lausanne, Switzerland. Diffractometer settings were 40 kV,
30 mA, 3-80 2, step scanning with 0.02 2 step size and
2 s counting time per step. The procedure for the identifica-
tion of the clay minerals was as described in Moore and
Reynolds (22). The <2 m fraction was separated by
centrifugation, and oriented samples were prepared on glass
slides. The samples were analyzed by XRD before and after
glycerol treatment, K- and Mg-saturation, and heat treatment
(550 C, 1h). The bulk geochemical compositions of the solid
samples were determined by X-ray fluorescence (XRF; Philips
PW 2400) for 38 elements. Fifteen selected solid samples
were subjected to a seven-step sequential extraction pro-
cedure taken from Dold (23) at the SGS Services Laboratory,
Toronto, Canada.
2.2.2. Acid-Potential (AP). Sequential extraction data can
be used as a high-resolution approach for the AP in acid-
base accounting. The sum of the sulfur concentrations in
the seven steps of the sequential extraction (measured by
ICP-AES) represents the total sulfur concentration of the
sample (Stotal). The sum of steps 1-4 represents the sulfate
concentration (Ssulfate) (23). The difference between Stotal and
Ssulfate represents Ssulfide, which was calculated as pyrite
equivalent, since pyrite was the dominant sulfide mineral at
Chanaral.
2.2.3. Aqueous Chemistry. The 29 pore water samples
were analyzed for major cations (Na+, K+, Mg2+, and Ca2+)
and anions (Cl-, NO2-, NO3-, and SO42-) by ion chroma-
tography (IC; Dionex DX 120) at the Centre dAnalyse
Minerale, University of Lausanne, Switzerland. Trace metals
VOL. 40, NO. 3, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 753
 TABLE 1: Acid Potential Calculations Based on Sulfur
Concentrations in the Sequential Extraction Leachesa
SAP
pyrite (sulfide acid
Stotal Ssulfate Ssulfide equivalent potential)
sample wt % wt % wt % wt % t CaCO3/1000t
CT2-10 1 0.71 0.29 0.54 18.13
CH2/S 4.29 3.99 0.30 0.56 18.75
CH2/2-5 2.06 1.33 0.73 1.36 45.63
CH2/40-60 1.18 0.63 0.55 1.02 34.38
CH2/70-73 1 0.69 0.31 0.58 19.38
CH2/90-100 0.58 0.18 0.40 0.74 25.00
CH2/150-170 0.19 0.03 0.16 0.30 10.00
CH2/280-300 0.44 0.03 0.41 0.76 25.63
CH3/1-2 1.32 0.83 0.49 0.91 30.63
FIGURE 2. Distribution of pH and Eh with depth, for piezometer CH3/03 1.64 1.12 0.52 0.97 32.50
nests CH1, CH2, and CH3. CH3/35-55 0.65 0.16 0.49 0.91 30.63
CH3/130-150 0.68 0.24 0.44 0.82 27.50
CH3/188-200 0.39 0.04 0.49 0.90 30.31
(Fe, Al, Cd, Cu, Mn, Zn, Pb, V, Cr, Co, Ni, As, Se, Sr, Mo, Ba, CH3/280-300 0.45 0.03 0.42 0.78 26.25
Tl, and U) were analyzed by inductively coupled plasma- mean 1.13 0.72 0.43 0.80 26.76
mass spectroscopy ICP-MS (PerkinElmer HP 4500 series SD 1.01 1.00 0.13 0.25 8.32
1000) at the Institute F.-A- Forel, University of Geneva, a The sum of the S concentrations in the seven steps (measured by
Switzerland. ICP-AES) represents the total sulfur concentration of the sample (Stotal).
The sum of step 1-4 represents the sulfate concentration (Ssulfate). The
3. Results and Discussion difference between Stotal and Ssulfate represents Ssulfide, which was
calculated as pyrite equivalent.
3.1 Stratigraphy and Mineralogy. On the surface of the
tailings, greenish-bluish and white colored efflorescent salts
were abundant. XRD analysis revealed that these were mainly Cu and Zn, showed a strong enrichment toward the surface,
eriochalcite (CuCl2H2O) and halite (NaCl), which are both indicating a greater mobility under the acid conditions of
water-soluble. The stratigraphy of the tailings was character- the oxidation zone. In contrast, Pb did not show this
ized by a low pH (2.6-4) oxidation zone, with a depth of 70 enrichment, possibly due to anglesite (PbSO4) formation,
cm at CH1, 73 cm at CH2, and 188 cm at CH3, and an which has lower solubility than Cu and Zn sulfates. Arsenic
underlying primary zone with a constant neutral pH (Figure showed a decrease from the primary zone (143-243 mg/kg)
2). CH3 and CH4 had predominantly sandy grain size tailings toward the oxidation zone (30-90 mg/kg) in CH1 and CH2.
material, while CH1 and CH2 had silty to clayey grain size Additionally, a decrease in concentration from CH1 (143-
material. The oxidation zone was characterized by secondary 243 mg/kg) toward CH4 (26-52 mg/kg) was observed. This
ferric mineralogy, which gives the typical yellowish-brown suggests that As was leached from the oxidation zone and
color (dominated by jarosite) with ochre to orange streaks washed out at the beach area by the tidal cycle, what is also
and dots of Fe(III) hydroxides. Gypsum was also present in supported by sequential extraction data (see below). Heavy
the oxidation zone. The clay mineralogy of the oxidation metals and As concentrations in the Chanaral tailings showed
zone was characterized by the presence of illite, kaolinite, clearly higher values (an average of one hundred times higher
and a vermiculite-type mixed-layer mineral. The latter in the case of copper) compared to the background con-
mineral typically results from biotite alteration in the centrations of a natural, uncontaminated beach in the area
oxidation zone of porphyry copper tailings (24). The peak (Playa Pan de Azucar; PPA) and to some international
position of the vermiculite-type mixed-layer at Chanaral at background concentrations of soils (Table 2).
11.8 suggested that K was largely replaced by Cu interlayer 3.3. Hydrogeochemistry and Sequential Extractions. The
cations, as observed at other tailings, where strong copper groundwater levels during sampling were at 1, 1.3, and 2 m
migration upward to the tailings surface was due to evapo- depths in CH1, CH2, CH3, respectively. The groundwater
ration-induced capillary transport (24). The gangue mineral- level varies more during the tidal cycle toward the sea (1
ogy of the primary zone was characterized by quartz, feldspar m at CH3) than toward the village (50 cm at CH1). The pore
(albite-anorthite), muscovite-biotite, kaolinite, magnetite, water pH was between 2.6 and 4 in the oxidation zone, and
and rutile. The main sulfide minerals were pyrite . chal- the pH increased sharply to neutral values below the oxidation
copyrite and chalcocite. The tailings contained on average front (Figure 2). At CH3, the pH was generally higher in the
0.8 ( 0.25 wt % pyrite equivalent, as calculated using the oxidation zone (around 4) than in CH1 and CH2 (around
sulfide sulfur concentrations from the sequential extractions 2.5), and it showed a peak of near neutral pH at a depth of
(Table 1). 60 cm, which was associated with an alluvial intercalation.
3.2 Bulk Geochemistry (XRF). The major elements, Al, The redox potential (Eh) showed values of up to 600 mV in
Si, and Ti, showed a homogeneous distribution in the primary the oxidation zone and decreased to 200 mV in the reducing
zone and the lower part of the oxidation zone (data not environment of the saturated zone (Figure 2). The alkalinity
shown). In the two uppermost samples, these elements was constant around 150 mg/L CaCO3 in the saturated zone.
decreased, whereas S, Cl, and Cu strongly increased (up to The distribution of the major anions (SO42- and Cl-) and
4.2, 15, and 2.4 wt %, respectively). This can be explained by cations (Na+, K+, Ca2+, and Mg2+) suggested that three
the strong capillary transport of Cu, SO4, and Cl toward the different types of waters were mixing in the tailings deposit
surface, and the subsequent dilution of the major element (Figure 3):
concentrations (Al, Si, and Ti) of the tailings material, as (1) High-saline waters (pH 7; up to 56 g/L Cl, 35 g/L Na,
reported for the El Salvador tailings impoundment no.1 (24). and 1900 mg/L SO4) were infiltrating the lower part of the
This was confirmed by the dominance of CuCl2H2O and tailings stratigraphy (below 4 m at CH1). These waters most
halite (NaCl) at the tailings surface. Heavy metals, such as likely originated from a subsurface river, which flows below

754 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 3, 2006

TABLE 2: Bulk Concentrations (XRF) of Relevant Heavy Metals and As in the Tailings of Chanaral Compared to Playa Pan De
Azucar and Concentrations in Reference Soilsa
Cu mg/kg Zn mg/kg Ni mg/kg Pb mg/kg Mo mg/kg As mg/kg
tailings Chanaral 1000-24 100 24-223 5-370 <3-26 19-186 30-281
mean tailings Chanaral 3500 52 102 10 72 111
Playa Pan de Azucar (PPA) 34 26 54 16 15 <3
background soils* 36 140-300 35 85-100 10 29
a
Data from Reimann and Caritat, 1998.

FIGURE 3. Hydrogeochemical data of major cations and anions

with depth for piezometer nests CH1, CH2, and CH3.
the surface along most of its riverbed in the El Salado valley
(The Salty). These waters have its source in the Salar of
Pedernales (150 km east of Chanaral). It is assumed that the
El Salado waters infiltrated the southern part of the tailings
disposal, supported by satellite images and the periodically
occurring alluviums.
(2) A plume of less-saline waters (pH 7; 0.1-0.9 g/L Cl,
0.2-0.8 g/L Na, and 400-900 mg/L SO4) waters was overlaying
the highly saline waters. These less-saline waters may
originate from leaks in the freshwater supply of Chanaral or
an unknown groundwater source or spring.
(3) These two waters mixed toward the sea with seawater
(pH 7.9; 17.8 g/L Cl, 11.5 g/L Na, and 2000 mg/L SO4), resulting
in intermediate concentrations of 13 g/L Cl and 11 g/L at
CH3 in the upper 2 m and of about 40 g/L Cl and 30 g/L Na
at CH2 below 3 m. All major cations and anions, except K,
showed a strong increase toward the surface in the unsatur-
ated zone due to evaporation-induced capillarity transport
FIGURE 4. Hydrogeochemical data of cations (Cu, Zn, Ni, Fe) and
oxyanions (Mo, As) with depth for piezometer nests CH1, CH2, and
CH3.
(Figure 3). K was retained in the low-pH oxidation zone by
the precipitation of jarosite. This was supported by step 4
(NH4OxH leach) in the sequential extractions, in which K
increased up to 0.14 wt % in the oxidation zone (Figure 5),
and jarosite dissolved completely (23).
The divalent metals concentrations greatly increased from
the pore water in the low-pH oxidation zone toward the
surface (up to 2265 mg/L Cu, up to 20 mg/L Zn, up to 18
mg/L Ni, and up to 183 mg/L Fe, respectively; Figure 4). This
was documented by a strong enrichment in Cu, up to 1.9 wt
% in the water-soluble fraction of the sequential extractions
(Figure 6). In the neutral pH-saturated zone, the mobility of
Cu and Zn was expected to be very low due to sorption.
However, high mobility of these divalent cations was observed
(up to 19 mg/L Cu and 12 mg/L Zn, respectively) in the high-

VOL. 40, NO. 3, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 755

FIGURE 5. Data for Fe, S, and K concentrations from sequential
extractions with depth for sampling points CH2, CH3, and CH4. CH4
is located directly at the beach. For better visibility of the data,
different scales are applied at the different sampling points.
Abbreviations: NH4-Ac, ammonium acetate; NH4-OxD, ammonium
oxalate, darkness; NH4-OxH, ammonium oxalate, 80 C; cv, covellite;
cc-dg, chalcocite-digenite.
saline waters (Table 2; Fig, 4). Geochemical modeling
(PHREEQC) suggested that these metals were transported as
chloro-complexes. These high Cu and Zn concentrations
could not have originated from the oxidation zone. The
underlying less-saline neutral waters act as a geochemical
FIGURE 6. Data for Cu, As, and Mo from sequential extractions with
depth for sampling points CH2, CH3, and CH4. CH4 is located directly
at the beach. For better visibility of the data, different scales are
applied at the different sampling points.
barrier for the downward mobility of the liberated Cu and
Zn from the oxidation zone. Also, the liberation of Cu and
Zn from the sulfide mineral assemblage at the prevailing
reducing (100-200 mV) conditions in the high-saline satu-
rated zone of the tailings is unlikely. Therefore, the source
of these high Cu and Zn concentrations in the neutral high-
saline waters has to be searched for in the upstream riverbed
of the El Salado valley.

TABLE 3: Concentrations (ICP-MS) of Relevant Heavy Metals and As in Solution, in Pore Water of the Tailings at Chanaral,
Compared to Background Values in Ocean Waters and Different Action Levels for Drinking Water Standardsa
Cu mg/L Zn mg/L Ni mg/L Pb mg/L Mo mg/L As mg/L
pore water tailings Chanaral max. concentrations 2265 20 18 0.046 <0.0002 <0.0002
in the oxidation zone (pH 2-4)
pore water tailings Chanaral max. concentrations 19 12 0.28 0.005 10 0.608
in the primary zone (pH 7)
action levels for drinking water* 0.3-2 0.3-5 0.02-0.1 0.1-0.3 0.07-0.25 0.01-0.07
background ocean water* 0.00025 0.005 0.00056 0.00003 0.01 0.0037
a Data from Reimann and Caritat, 1998.

756 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 3, 2006


FIGURE 7. Schematic model of element flows at the Chanaral marine
shore tailings deposit (for explanation see text).
The elements, Mo and As, which are stable as oxyanions,
were below detection limit in the oxidation zone, most likely
due to their adsorption onto the secondary Fe(III) hydroxides
at low pH conditions (24). Sequential extraction data
confirmed that As and Mo were mainly associated with the
Fe(III) hydroxide and oxide fractions in the low-pH oxidation
zone (Figure 6). In the saturated zone, these elements showed
high concentrations in the pore waters (up to 608 g/L As
and up to 10266 g/L Mo; Figure 4), where higher pH and
more reducing condition increased their mobility. This was
supported by the sequential extraction data, which showed
As present below the oxidation front mainly in the exchange-
able fraction and associated with the sulfide fraction (Figure
6). The relevant heavy metals and As concentrations detected
in the pore water of the tailings of Chanaral are strongly
increased in comparison to the background concentrations
of ocean water and to some international action levels for
drinking water standards (Table 3).
The mineralogical and geochemical study of the shore
mine tailings deposit at Chanaral, Chile, allows for the
development of a schematic element flow model associated
with this system (Figure 7). The data indicate that two element
flow directions (toward the tailings surface and toward the
sea) and two contaminant groups with different geochemical
behaviors (cations as Cu2+, Zn2+, Ni2+, and oxyanions as
H2AsO42- and HMoO4-) can be distinguished. The geochem-
ical data suggest that the hydrology in the studied southern
part of the tailings system was primarily controlled by the
tidal cycle, by the infiltration of the high-saline, Cu, and Zn-
rich waters from the El Salado valley, and by an unknown
source of less-saline waters.
The divalent metal cations, Cu2+, Ni2+, and Zn2+, liberated
in the oxidation zone, were mobile under acid conditions
and transported to the tailings surface via capillarity; where
supersaturation led to the precipitation of these metals as
secondary metal chlorides or sulfates (e.g. eriochalcite).
Subsequently, these mainly water-soluble secondary minerals
became exposed to eolian transport by dominated W-SW
winds toward the Village of Chanaral.
In contrast, As and Mo did not show any mobilization
toward the surface of the tailings. This was due to their strong
adsorption to secondary Fe(III) hydroxides in the low-pH
oxidation zone. The residual concentrations of As at the
tailings surface, which was exposed to eolian transport, were
between 14 and 61 mg/kg. The data suggested that an
important part of As and Mo was leached out toward the sea.
It can be postulated that the sea is mainly affected by As,
Mo, Cu, and Zn contamination by infiltration and colloidal
transport, while the population of Chanaral is mainly exposed
to increased concentrations of Cu. Ni, and Zn in water-
soluble form via eolian transport.
Acknowledgments
I thank the Mayor of the Village of Chanaral, Mr. H. Volta,
and his staff for the permission and support in performing
this study. Special thanks are due to Miguel and Mihailo
Nenadovich and the Sociedad Minera Legal C-Uno de la
Bahia de Chanaral for their interest, support, and permission
for this study, and for access to their properties. I also thank
J. Palma, P. Riveros, and O. Monroy for the supply of
important information and their support; the Management
of CODELCO, Division El Salvador, J. Cifuentes, E. Silva, R.
Vargas, and C. Rodriguez for their support and interest in
this study, and finally, N. Diaby, M. Perret, J.-C. Lavanchy,
J.-P. Favarger, L. Dufresne, P. Thelin, and H. R. Pfeifer for
their help with the fieldwork, sample preparation, analysis,
and support. The study was funded by the Swiss National
Science Foundation project no. 200021-105507/1.
Literature Cited
(1) Burd, B.; Macdonald, R.; Boyd, J. Punctuated recovery of
sediments and benthic infauna: a 19-year study of tailings
deposition in a British Columbia fjord. Mar. Environ. Res. 2000,
49, 145-175.
(2) Castilla, J. C. Environmental impact in sandy beaches of copper
mine tailings at Chanaral, Chile. Mar. Pollut. Bull. 1983, 14,
459-464.
(3) David, C. P. Heavy metal concentrations in growth bands of
corals: a record of mine tailings input through time (Marinduque
Island, Philippines). Mar. Pollut. Bull. 2003, 46, 187-196.
(4) Pedersen, T. F.; Mueller, B.; McNee, J. J. The early diagenesis
of submerged sulphide-rich mine tailings in Anderson Lake,
Manitoba. Can. J. Earth Sci. 1993, 30, 1099-1109.
(5) Elberling, B.; Luise Knudsen, K.; Kristensen, P. H.; Asmund, G.
Applying foraminiferal stratigraphy as a biomarker for heavy
metal contamination and mining impact in a fjord in West
Greenland. Mar. Environ. Res. 2003, 55, 235-256.
(6) Olsgard, F.; Hasle, J. R. Impact of waste from titanium mining
on benthic fauna. J. Exp. Mar. Biol. Ecol. 1993, 172, 185-213.
(7) Lee, M. R.; Correa, J. A. Effects of copper mine tailings disposal
on littoral meiofaunal assemblages in the Atacama region of
northern Chile. Mar. Environ. Res. 2005, 59, 1-18.
(8) Castilla, J. C.; Nealler, E. Marine environmental impact due to
mining activities of El Salvador copper mine, Chile. Mar. Pollut.
Bull. 1978, 9, 67-70.
(9) Farina, J. M.; Castilla, J. C. Temporal variation in the diversity
and cover of sessile species in rocky intertidal communities
affected by copper mine tailings in northern Chile. Mar. Pollut.
Bull. 2001, 42, 554-568.
(10) Lee, M. R.; Correa, J. A.; Castilla, J. C. An assessment of the
potential use of the nematode to copepod ratio in the monitoring
of metals pollution. The Chanaral case. Mar. Pollut. Bull. 2001,
42, 696-701.
(11) Lee, M. R.; Correa, J. A.; Zhang, H. Effective metal concentrations
in porewater and seawater labile metal concentrations associ-
ated with copper mine tailings disposal into the coastal waters
of the Atacama region of northern Chile. Mar. Pollut. Bull. 2002,
44, 956-961.
(12) Medina, M.; Andrade, S.; Faugeron, S.; Lagos, N.; Mella, D.;
Correa, J. A. Biodiversity of rocky intertidal benthic communities
associated with copper mine tailing discharges in northern Chile.
Mar. Pollut. Bull. 2005, 50, 396-409.
(13) Ramirez, M.; Massolo, S.; Frache, R.; Correa, J. A. Metal speciation
and environmental impact on sandy beaches due to El Salvador
copper mine, Chile. Mar. Pollut. Bull. 2005, 50, 62-72.
(14) Zhang, H.; Zhao, F.-J.; Sun, B.; Davidson, W.; McGrath, S. P. A
new method to mesure effective soil solution concentration
predicts copper availability to plants. Environ. Sci. Technol. 2001,
35, 2602-2607.
(15) Camus, F. Geologa de los sistemas porfricos en los Andes de
Chile; Servicio Nacional de Geologa y Minera: Santiago de
Chile, 2003.
(16) Gustafson, L. B.; Hunt, J. P. The porphyry copper deposit at El
Salvador, Chile. Econ. Geol. 1975, 70, 857-912.
(17) Gustafson, L. B.; Quiroga, J. Patterns of mineralization and
alteration below the porphyry copper orebody at El Salvador,
Chile. Econ. Geol. 1995, 90, 2-16.
(18) Larrain, H.; Velasquez, F.; Cereceda, P.; Espejo, R.; Pinto, R.;
Osses, P.; Schemenauer, R. S. Fog measurements at the site
Falda Verde north of Chanaral compared to other fog stations
of Chile. Atmos. Res. 2002, 64, 273-284.
(19) Patterson, R. J.; Frape, S. K.; Dykes, L. S.; McLeod, R. A. A coring
and sqeezing technique for detailed study of subsurface water
chemistry. Can. J. Earth Sci. 1978, 15, 162-169.
(20) Nordstrom, D. K. Thermochemical redox equilibria of Zobells
solution. Geochim. Cosmochim. Acta 1977, 41, 1835-1841.
(21) Light, T. S. Standard solution for redox potential measurement.
Anal. Chem. 1972, 44, 1038-1039.
VOL. 40, NO. 3, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 757
(22) Moore, D. M.; Reynolds, R. C. X-ray diffraction and the
identification and analysis of clay minerals, 2nd ed.; Oxford
University Press: Oxford, 1997.
(23) Dold, B. Speciation of the most soluble phases in a sequential
extraction procedure adapted for geochemical studies of copper
sulfide mine waste. J. Geochem. Explor. 2003, 80, 55-68.
(24) Dold, B.; Fontbote, L. Element cycling and secondary mineralogy
in porphyry copper tailings as a function of climate, primary
mineralogy, and mineral processing. J. Geochem. Explor. 2001,
74, 3-55.
Received for review July 28, 2005. Revised manuscript re-
ceived November 3, 2005. Accepted November 16, 2005.
ES051475Z

758 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 3, 2006